Project-MONAI / Project-MONAI/MONAI
OptionalImportError when using pytorch 2.7
Open
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 8.7k
- Forks
- 1.6k
- Avg merge
- 5d 1h
- Merged PRs (30d)
- 20
Description
22:12:13 ======================================================================
22:12:13 ERROR: test_milmodel (unittest.loader._FailedTest)
22:12:13 ----------------------------------------------------------------------
22:12:13 ImportError: Failed to import test module: test_milmodel
22:12:13 Traceback (most recent call last):
22:12:13 File "/usr/lib/python3.10/unittest/loader.py", line 154, in loadTestsFromName
22:12:13 module = __import__(module_name)
22:12:13 File "/tmp/tmp.tAhlxKhK20/tests/networks/nets/test_milmodel.py", line 53, in <module>
22:12:13 backbone = models.densenet121(pretrained=False)
22:12:13 File "/usr/local/lib/python3.10/dist-packages/monai/utils/module.py", line 420, in __getattr__
22:12:13 raise self._exception
22:12:13 File "/usr/local/lib/python3.10/dist-packages/monai/utils/module.py", line 378, in optional_import
22:12:13 pkg = __import__(module) # top level module
22:12:13 File "/usr/local/lib/python3.10/dist-packages/torchvision/__init__.py", line 10, in <module>
22:12:13 from torchvision import _meta_registrations, datasets, io, models, ops, transforms, utils # usort:skip
22:12:13 File "/usr/local/lib/python3.10/dist-packages/torchvision/_meta_registrations.py", line 26, in <module>
22:12:13 def meta_roi_align(input, rois, spatial_scale, pooled_height, pooled_width, sampling_ratio, aligned):
22:12:13 File "/usr/local/lib/python3.10/dist-packages/torchvision/_meta_registrations.py", line 18, in wrapper
22:12:13 if torchvision.extension._has_ops():
22:12:13 monai.utils.module.OptionalImportError: import torchvision.models (partially initialized module 'torchvision' has no attribute 'extension' (most likely due to a circular import)).
22:12:13
22:12:13 For details about installing the optional dependencies, please visit:
22:12:13 https://docs.monai.io/en/latest/installation.html#installing-the-recommended-dependencies
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reproducing the failure from tests/networks/nets/test_milmodel.py, especially the models.densenet121(pretrained=False) import, with the reported PyTorch 2.7 environment. Inspect the OptionalImportError traceback through monai/utils/module.py and verify the test module imports successfully with the relevant optional dependencies installed.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning, testing
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100