pytorch / pytorch/vision

Nightly build flaky pytorch/vision / conda-py3_11-cpu builds

Open
#8,372 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
17.9k
Forks
7.3k
Avg merge
1d 15h
Merged PRs (30d)
13

Description

🐛 Describe the bug

Flaky issue on pytorch/vision / conda-py3_11-cpu builds. Has been happening for a while now.
Most likely due to corrupt worker environment:

+ __conda_exe run -p /Users/ec2-user/runner/_work/_temp/pytorch_pkg_helpers_8521283920_smoke python3 pytorch/vision/test/smoke_test.py
+ /opt/homebrew/Caskroom/miniconda/base/bin/conda run -p /Users/ec2-user/runner/_work/_temp/pytorch_pkg_helpers_8521283920_smoke python3 pytorch/vision/test/smoke_test.py
/Users/ec2-user/.local/lib/python3.11/site-packages/torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: ''If you don't plan on using image functionality from `torchvision.io`, you can ignore this warning. Otherwise, there might be something wrong with your environment. Did you have `libjpeg` or `libpng` installed before building `torchvision` from source?
  warn(
torchvision: 0.19.0a0+480eec2
Traceback (most recent call last):
  File "/Users/ec2-user/runner/_work/vision/vision/pytorch/vision/test/smoke_test.py", line 103, in <module>
    main()
  File "/Users/ec2-user/runner/_work/vision/vision/pytorch/vision/test/smoke_test.py", line 83, in main
    print(f"{torch.ops.image._jpeg_version() = }")
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/ec2-user/runner/_work/_temp/pytorch_pkg_helpers_8521283920_smoke/lib/python3.11/site-packages/torch/_ops.py", line 927, in __getattr__
    raise AttributeError(
AttributeError: '_OpNamespace' 'image' object has no attribute '_jpeg_version'

ERROR conda.cli.main_run:execute(124): `conda run python3 pytorch/vision/test/smoke_test.py` failed. (See above for error)
torch.cuda.is_available: False

Rerun is ususally successful

Versions

0.19.0

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with pytorch/vision/test/smoke_test.py, especially the failing _jpeg_version call, and reproduce the conda-py3_11-cpu nightly build or rerun. Compare the worker environment and image-extension setup between failed and successful runs. Done means the flaky smoke build is diagnosed and the nightly build no longer fails intermittently.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
ci-cd, machine-learning, testing-qa
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.