FileNotFoundError in test_perceptual_loss

[KumoLiu]

[2024-01-10T02:16:36.089Z] ======================================================================
[2024-01-10T02:16:36.089Z] ERROR: test_identical_input_2 (tests.test_perceptual_loss.TestPerceptualLoss)
[2024-01-10T02:16:36.089Z] ----------------------------------------------------------------------
[2024-01-10T02:16:36.089Z] Traceback (most recent call last):
[2024-01-10T02:16:36.089Z]   File "/usr/local/lib/python3.10/dist-packages/parameterized/parameterized.py", line 620, in standalone_func
[2024-01-10T02:16:36.089Z]     return func(*(a + p.args), **p.kwargs, **kw)
[2024-01-10T02:16:36.089Z]   File "/opt/monai/tests/test_perceptual_loss.py", line 65, in test_identical_input
[2024-01-10T02:16:36.089Z]     loss = PerceptualLoss(**input_param)
[2024-01-10T02:16:36.089Z]   File "/opt/monai/monai/losses/perceptual.py", line 107, in __init__
[2024-01-10T02:16:36.089Z]     self.perceptual_function = RadImageNetPerceptualSimilarity(net=network_type, verbose=False)
[2024-01-10T02:16:36.090Z]   File "/opt/monai/monai/losses/perceptual.py", line 257, in __init__
[2024-01-10T02:16:36.090Z]     self.model = torch.hub.load("Warvito/radimagenet-models", model=net, verbose=verbose)
[2024-01-10T02:16:36.090Z]   File "/usr/local/lib/python3.10/dist-packages/torch/hub.py", line 565, in load
[2024-01-10T02:16:36.090Z]     model = _load_local(repo_or_dir, model, *args, **kwargs)
[2024-01-10T02:16:36.090Z]   File "/usr/local/lib/python3.10/dist-packages/torch/hub.py", line 594, in _load_local
[2024-01-10T02:16:36.090Z]     model = entry(*args, **kwargs)
[2024-01-10T02:16:36.090Z]   File "/root/.cache/torch/hub/Warvito_radimagenet-models_main/radimagenet_models/models/resnet.py", line 153, in radimagenet_resnet50
[2024-01-10T02:16:36.090Z]     model.load_state_dict(torch.load(cached_file))
[2024-01-10T02:16:36.090Z]   File "/usr/local/lib/python3.10/dist-packages/torch/serialization.py", line 988, in load
[2024-01-10T02:16:36.090Z]     with _open_file_like(f, 'rb') as opened_file:
[2024-01-10T02:16:36.090Z]   File "/usr/local/lib/python3.10/dist-packages/torch/serialization.py", line 437, in _open_file_like
[2024-01-10T02:16:36.090Z]     return _open_file(name_or_buffer, mode)
[2024-01-10T02:16:36.090Z]   File "/usr/local/lib/python3.10/dist-packages/torch/serialization.py", line 418, in __init__
[2024-01-10T02:16:36.090Z]     super().__init__(open(name, mode))
[2024-01-10T02:16:36.090Z] FileNotFoundError: [Errno 2] No such file or directory: '/root/.cache/torch/hub/checkpoints/RadImageNet-ResNet50_notop.pth'
[2024-01-10T02:16:36.090Z] 
[2024-01-10T02:16:36.090Z] ----------------------------------------------------------------------
[2024-01-10T02:16:36.090Z] Ran 13397 tests in 4093.907s
[2024-01-10T02:16:36.090Z] 
[2024-01-10T02:16:36.090Z] FAILED (errors=1, skipped=158)

[KumoLiu]

Access denied with the following error:

        Cannot retrieve the public link of the file. You may need to change
        the permission to 'Anyone with the link', or have had many accesses. 

The cause of the error appears to be that the permission of the link has changed.

MONAI/monai/losses/perceptual.py

Line 257 in e1ffa7e

self.model = torch.hub.load("Warvito/radimagenet-models", model=net, verbose=verbose)

Hi @marksgraham and @Warvito, would you mind taking a look at this one? Is there any reason you have done this or has it been done accidentally? Thanks!

[KumoLiu]

May not be due to permission but the issue described here.

[marksgraham]

I can now download so I think the problem must be hitting a daily download limit. I think we need to move to a non g-drive solution for hosting these files. How is this handled in the rest of MONAI?
@ericspod

[KumoLiu]

It may be about a daily download limit.
It has failed for three days and happened in other gdown-related tests.
#7383
Project-MONAI/tutorials#1608 (comment)

We can put some small files here for test.
https://github.com/Project-MONAI/MONAI-extra-test-data/releases

[ericspod]

This issue is either sorted with the gdown fix or it's a matter of moving temp files to the extra test data repo so I'd close this for now.