Random shuffle before dropping the last few samples#123
Open
tongzhoumu wants to merge 1 commit intoNVlabs:mainfrom
Open
Random shuffle before dropping the last few samples#123tongzhoumu wants to merge 1 commit intoNVlabs:mainfrom
tongzhoumu wants to merge 1 commit intoNVlabs:mainfrom
Conversation
tongzhoumu
force-pushed
the
drop-last
branch
from
d02058f to
28832de
Compare
gheinrich
pushed a commit
to gheinrich/VILA
that referenced
this pull request
Dec 16, 2024
* Add files via upload * Add files via upload * Add files via upload * Rename eval_benchmark_1_correctness.py to eval_benchmark_correctness.py * Rename eval_benchmark_2_detailed_orientation.py to eval_benchmark_detailed_orientation.py * Rename eval_benchmark_3_context.py to eval_benchmark_context.py * Rename eval_benchmark_4_temporal.py to eval_benchmark_temporal.py * Rename eval_benchmark_5_consistency.py to eval_benchmark_consistency.py * Update eval_all.sh * Update eval_all.sh * Update eval_all.sh * Create run_vila_benchmark.sh * Update slurm_run_all.sh * Update eval_all.sh * Update eval_all.sh * Create eval_vila_benchmark_gpt4.sh * Update eval_vila_benchmark_gpt4.sh * Update eval_all.sh * Update eval_all.sh * Delete llava/eval/video/convert_pred_to_json.py * Update model_vqa_videodemo_benchmark.py * Update eval_all.sh * Update run_vila_benchmark.sh * Update slurm_run_all_sharded.sh * Update slurm_run_all.sh
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
I noticed a bug in the data sampler. In the original implementation, the same elements will be dropped in every epoch.
For example, assume the dataset size is 900, and the batch size is 200, then the same 100 samples are dropped every time, which means they are never used in training