-
Notifications
You must be signed in to change notification settings - Fork 3.4k
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
Switch to PyTorch 1.6 in Drone CI #4393
Conversation
Codecov Report
@@ Coverage Diff @@
## master #4393 +/- ##
======================================
Coverage 92% 93%
======================================
Files 116 116
Lines 8716 8717 +1
======================================
+ Hits 8060 8082 +22
+ Misses 656 635 -21 |
1.7 is causing the failed tests. |
GPU tests are failing... |
The failing amp test is |
Hey @ydcjeff, sorry for creating a bug. I am on it :) Best regards, |
* switch to 1.6 * readme * 1.7 * back to normal [ci skip] * horovodrun --verbose * try with apex * add apex test * change base * description * test with 1.7 * back to 1.6 * no gradient_clip_val * re-add gradient_clip_val * no amp * temp skip torch.cuda.amp + horovod test * Apply suggestion from code review Co-authored-by: Sean Naren <sean.narenthiran@gmail.com> * Fix formatting * ddp * Moved extended model outside of function to prevent pickling issue for drone * typo * resolve bug * extract automatic_automization Co-authored-by: Sean Naren <sean.narenthiran@gmail.com> Co-authored-by: SeanNaren <sean@grid.ai> Co-authored-by: chaton <thomas@grid.ai> (cherry picked from commit ee414d2)
* switch to 1.6 * readme * 1.7 * back to normal [ci skip] * horovodrun --verbose * try with apex * add apex test * change base * description * test with 1.7 * back to 1.6 * no gradient_clip_val * re-add gradient_clip_val * no amp * temp skip torch.cuda.amp + horovod test * Apply suggestion from code review Co-authored-by: Sean Naren <sean.narenthiran@gmail.com> * Fix formatting * ddp * Moved extended model outside of function to prevent pickling issue for drone * typo * resolve bug * extract automatic_automization Co-authored-by: Sean Naren <sean.narenthiran@gmail.com> Co-authored-by: SeanNaren <sean@grid.ai> Co-authored-by: chaton <thomas@grid.ai> (cherry picked from commit ee414d2)
* switch to 1.6 * readme * 1.7 * back to normal [ci skip] * horovodrun --verbose * try with apex * add apex test * change base * description * test with 1.7 * back to 1.6 * no gradient_clip_val * re-add gradient_clip_val * no amp * temp skip torch.cuda.amp + horovod test * Apply suggestion from code review Co-authored-by: Sean Naren <sean.narenthiran@gmail.com> * Fix formatting * ddp * Moved extended model outside of function to prevent pickling issue for drone * typo * resolve bug * extract automatic_automization Co-authored-by: Sean Naren <sean.narenthiran@gmail.com> Co-authored-by: SeanNaren <sean@grid.ai> Co-authored-by: chaton <thomas@grid.ai>
What does this PR do?
Follow up of #3658
Fixes #4484
Related: https://app.circleci.com/pipelines/github/pytorch/pytorch-integration-testing/74/workflows/4a7bc96a-c550-4f0e-8c51-eb9d9af6166c/jobs/451/tests
PR review
Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in Github issues there's a high chance it will not be merged.
Did you have fun?
Make sure you had fun coding 🙃