-
Notifications
You must be signed in to change notification settings - Fork 352
[BE] Fix MPS experimental workflow #2181
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
By actually running them in the conda env that is being created
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/2181
Note: Links to docs will display an error until the docs builds have been completed. ⏳ No Failures, 6 PendingAs of commit f1d2c8b with merge base c9b9adc ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
does this PR break this? https://github.com/pytorch/ao/actions/runs/14888035472/job/41812796233 |
|
That's extremely unlikely since this a workflow only change to an unrelated workflow. Considering you're linking to a failed nightly job try pinning the torch version to see if that fixes |
|
cc oncall @andrewor14 |
This is fixed separately in #2188 |
* [BE] Fix MPS tests By actually running them in the conda env that is being created * Update torchao_experimental_test.yml * Update torchao_experimental_test.yml
* [BE] Fix MPS tests By actually running them in the conda env that is being created * Update torchao_experimental_test.yml * Update torchao_experimental_test.yml
By actually running them in the conda env that is being created, rather than rely on system-wide Python
Otherwise, as one can see from https://github.com/pytorch/ao/actions/runs/14861926861/job/41728802819
pip installare not running in the conda environment that were configured for the op, but rather against default Python version installed on the system, which has recently changed to 3.13