Error while running DLMUSE individually on the cluster

[euroso97]

Reporting the following errors. I used 'pip install DLMUSE'.

Case #1: Initial run (including the model downloading process)

Renaming dic is saved to /Projects/mask_sample_DLMUSE/renamed_image/renaming.json
Using model folder: /cbica/home/baikk/.conda/envs/NIB/lib/python3.12/site-packages/DLMUSE/nnunet_results/Dataset903_Task903_DLMUSEV2/nnUNetTrainer__nnUNetPlans__3d_fullres/
DLMUSE model not found, downloading...

Fetching 15 files:   0%|          | 0/15 [00:00<?, ?it/s]
Fetching 15 files:   7%|▋         | 1/15 [00:00<00:01,  7.81it/s]
Fetching 15 files:  53%|█████▎    | 8/15 [00:00<00:00, 13.08it/s]
Fetching 15 files:  73%|███████▎  | 11/15 [00:02<00:01,  3.40it/s]
Fetching 15 files:  80%|████████  | 12/15 [00:03<00:00,  3.30it/s]
Fetching 15 files: 100%|██████████| 15/15 [00:03<00:00,  4.94it/s]
DLMUSE model has been successfully downloaded!
Running in CUDA mode.
Traceback (most recent call last):
  File "/cbica/home/baikk/.conda/envs/NIB/bin/DLMUSE", line 8, in <module>
    sys.exit(main())
             ^^^^^^
  File "/cbica/home/baikk/.conda/envs/NIB/lib/python3.12/site-packages/DLMUSE/__main__.py", line 330, in main
    predictor.initialize_from_trained_model_folder(
  File "/cbica/home/baikk/.conda/envs/NIB/lib/python3.12/site-packages/nnunetv2/inference/predict_from_raw_data.py", line 84, in initialize_from_trained_model_folder
    checkpoint = torch.load(join(model_training_output_dir, f'fold_{f}', checkpoint_name),
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/cbica/home/baikk/.conda/envs/NIB/lib/python3.12/site-packages/torch/serialization.py", line 1470, in load
    raise pickle.UnpicklingError(_get_wo_message(str(e))) from None
_pickle.UnpicklingError: Weights only load failed. This file can still be loaded, to do so you have two options, �[1mdo those steps only if you trust the source of the checkpoint�[0m. 
	(1) In PyTorch 2.6, we changed the default value of the `weights_only` argument in `torch.load` from `False` to `True`. Re-running `torch.load` with `weights_only` set to `False` will likely succeed, but it can result in arbitrary code execution. Do it only if you got the file from a trusted source.
	(2) Alternatively, to load with `weights_only=True` please check the recommended steps in the following error message.
	WeightsUnpickler error: Unsupported global: GLOBAL numpy.core.multiarray.scalar was not an allowed global by default. Please use `torch.serialization.add_safe_globals([scalar])` or the `torch.serialization.safe_globals([scalar])` context manager to allowlist this global if you trust this class/function.

Check the documentation of torch.load to learn more about types accepted by default with weights_only https://pytorch.org/docs/stable/generated/torch.load.html.

Case #2: Second launch (the model was downloaded in the first step, so now skips the downloading and goes straight to running)

Renaming dic is saved to /Projects/renamed_image/renaming.json
Using model folder: /cbica/home/baikk/.conda/envs/NIB/lib/python3.12/site-packages/DLMUSE/nnunet_results/Dataset903_Task903_DLMUSEV2/nnUNetTrainer__nnUNetPlans__3d_fullres/
Loading the model...
Running in CUDA mode.
Traceback (most recent call last):
  File "/cbica/home/baikk/.conda/envs/NIB/bin/DLMUSE", line 8, in <module>
    sys.exit(main())
             ^^^^^^
  File "/cbica/home/baikk/.conda/envs/NIB/lib/python3.12/site-packages/DLMUSE/__main__.py", line 330, in main
    predictor.initialize_from_trained_model_folder(
  File "/cbica/home/baikk/.conda/envs/NIB/lib/python3.12/site-packages/nnunetv2/inference/predict_from_raw_data.py", line 84, in initialize_from_trained_model_folder
    checkpoint = torch.load(join(model_training_output_dir, f'fold_{f}', checkpoint_name),
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/cbica/home/baikk/.conda/envs/NIB/lib/python3.12/site-packages/torch/serialization.py", line 1470, in load
    raise pickle.UnpicklingError(_get_wo_message(str(e))) from None
_pickle.UnpicklingError: Weights only load failed. This file can still be loaded, to do so you have two options, �[1mdo those steps only if you trust the source of the checkpoint�[0m. 
	(1) In PyTorch 2.6, we changed the default value of the `weights_only` argument in `torch.load` from `False` to `True`. Re-running `torch.load` with `weights_only` set to `False` will likely succeed, but it can result in arbitrary code execution. Do it only if you got the file from a trusted source.
	(2) Alternatively, to load with `weights_only=True` please check the recommended steps in the following error message.
	WeightsUnpickler error: Unsupported global: GLOBAL numpy.core.multiarray.scalar was not an allowed global by default. Please use `torch.serialization.add_safe_globals([scalar])` or the `torch.serialization.safe_globals([scalar])` context manager to allowlist this global if you trust this class/function.

Check the documentation of torch.load to learn more about types accepted by default with weights_only https://pytorch.org/docs/stable/generated/torch.load.html.

[euroso97]

Seems to work fine when I used git clone installation method with "pip install -e ."