Thank you for your work. I recently ran the Scannet dataset and found the following problems.

[llllcs]

Traceback (most recent call last):
File "tools/train.py", line 136, in
main()
File "tools/train.py", line 132, in main
runner.train()
File "/root/miniconda3/lib/python3.8/site-packages/mmengine/runner/runner.py", line 1777, in train
model = self.train_loop.run() # type: ignore
File "/root/miniconda3/lib/python3.8/site-packages/mmengine/runner/loops.py", line 98, in run
self.run_epoch()
File "/root/miniconda3/lib/python3.8/site-packages/mmengine/runner/loops.py", line 115, in run_epoch
self.run_iter(idx, data_batch)
File "/root/miniconda3/lib/python3.8/site-packages/mmengine/runner/loops.py", line 131, in run_iter
outputs = self.runner.model.train_step(
File "/root/miniconda3/lib/python3.8/site-packages/mmengine/model/base_model/base_model.py", line 114, in train_step
losses = self._run_forward(data, mode='loss') # type: ignore
File "/root/miniconda3/lib/python3.8/site-packages/mmengine/model/base_model/base_model.py", line 361, in _run_forward
results = self(**data, mode=mode)
File "/root/miniconda3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/root/autodl-tmp/mmdetection3d-1.4.0/mmdet3d/models/detectors/base.py", line 75, in forward
return self.loss(inputs, data_samples, **kwargs)
File "/root/autodl-tmp/mmdetection3d-1.4.0/projects/oneformer3d/oneformer3d/oneformer3d.py", line 372, in loss
loss = self.criterion(x, sp_gt_instances)
File "/root/autodl-tmp/mmdetection3d-1.4.0/projects/oneformer3d/oneformer3d/unified_criterion.py", line 62, in call
loss = self.inst_criterion(pred, inst_gts)
File "/root/autodl-tmp/mmdetection3d-1.4.0/projects/oneformer3d/oneformer3d/instance_criterion.py", line 260, in call
indices.append(self.matcher(pred_instances, gt_instances))
File "/root/miniconda3/lib/python3.8/site-packages/torch/autograd/grad_mode.py", line 27, in decorate_context
return func(*args, **kwargs)
File "/root/autodl-tmp/mmdetection3d-1.4.0/projects/oneformer3d/oneformer3d/instance_criterion.py", line 500, in call
cost_value = torch.where(
RuntimeError: expected scalar type float but found double

log.txt

[filaPro]

Can you try with python package versions closer to the ones recommended in our Dockerfile? Including pytorch 1.11 -> 1.13.

[llllcs]

Thank you for your response. My server is unable to use the mirror file due to certain reasons. So, I created a virtual environment based on the dependency package versions from the mirror file you provided. However, during runtime, I encountered the following issue:
(mmdet3d) root@autodl-container-c35b448a49-d814657a:~/autodl-tmp/oneformer3d# python tools/train.py
configs/oneformer3d_1xb2_s3dis-area-5.py
Segmentation fault (core dumped)
After debugging, I found that the issue was caused by the following line in the configuration file:
custom_imports = dict(imports=['oneformer3d'])
Commenting out this line resulted in normal error reporting instead of a core dump. Do you have any solutions for this?

[filaPro]

Unfortunately, removing this line is not the solutions, as we need all imports from './oneformer3d' directory. I can only recommend to use Docker or virtualenv with exactly the same version of packages as in our Dockerfile.