-
Notifications
You must be signed in to change notification settings - Fork 25.7k
Uninstall PyTorch after testing on non-ephemeral Windows runners #97285
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/97285
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 67a287d: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
…285) Per title, I suspect that having a leftover PyTorch built from CUDA 11.7 installed in non-ephemeral Windows runners could cause some flakiness on Windows CUDA 11.8 jobs also running on the same type of runners, for example `win-vs2019-cuda11.8-py3` in https://hud.pytorch.org/pytorch/pytorch/commit/5d3c347bf6f0b86c96a1fe541db5d4f9586c8840 failed with a PATH error: ``` nvrtc: error: failed to open nvrtc-builtins64_117.dll. Make sure that nvrtc-builtins64_117.dll is installed correctly. ``` This also cleans up the dead code about `pytorch_env_restore.bat` under `ci_scripts` temp directory. This directory is cleaned up always by [teardown-win](https://github.com/pytorch/pytorch/blob/master/.github/actions/teardown-win/action.yml#L33). So the bat script will never be there for the next job anyway. As Windows test jobs are doing fine, proving that we don't need this adhoc script anymore. ### Testing https://github.com/pytorch/pytorch/actions/runs/4485931686/jobs/7888513795 Pull Request resolved: pytorch/pytorch#97285 Approved by: https://github.com/seemethere
…285) Per title, I suspect that having a leftover PyTorch built from CUDA 11.7 installed in non-ephemeral Windows runners could cause some flakiness on Windows CUDA 11.8 jobs also running on the same type of runners, for example `win-vs2019-cuda11.8-py3` in https://hud.pytorch.org/pytorch/pytorch/commit/5d3c347bf6f0b86c96a1fe541db5d4f9586c8840 failed with a PATH error: ``` nvrtc: error: failed to open nvrtc-builtins64_117.dll. Make sure that nvrtc-builtins64_117.dll is installed correctly. ``` This also cleans up the dead code about `pytorch_env_restore.bat` under `ci_scripts` temp directory. This directory is cleaned up always by [teardown-win](https://github.com/pytorch/pytorch/blob/master/.github/actions/teardown-win/action.yml#L33). So the bat script will never be there for the next job anyway. As Windows test jobs are doing fine, proving that we don't need this adhoc script anymore. ### Testing https://github.com/pytorch/pytorch/actions/runs/4485931686/jobs/7888513795 Pull Request resolved: pytorch/pytorch#97285 Approved by: https://github.com/seemethere
Per title, I suspect that having a leftover PyTorch built from CUDA 11.7 installed in non-ephemeral Windows runners could cause some flakiness on Windows CUDA 11.8 jobs also running on the same type of runners, for example
win-vs2019-cuda11.8-py3in https://hud.pytorch.org/pytorch/pytorch/commit/5d3c347bf6f0b86c96a1fe541db5d4f9586c8840 failed with a PATH error:This also cleans up the dead code about
pytorch_env_restore.batunderci_scriptstemp directory. This directory is cleaned up always by teardown-win. So the bat script will never be there for the next job anyway. As Windows test jobs are doing fine, proving that we don't need this adhoc script anymore.Testing
https://github.com/pytorch/pytorch/actions/runs/4485931686/jobs/7888513795