-
Notifications
You must be signed in to change notification settings - Fork 25.2k
Add FakeCrossRef tests for backwards, Fix Layer Norm Backward Decomp #85417
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
[ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/85417
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 471ddbc: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
…ard Decomp" [ghstack-poisoned]
test/test_ops.py
Outdated
aten.is_same_size.default, | ||
|
||
) | ||
# TODO: enable check_aliasing, too many failures :/ |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
ALL of the primtorch metas alias incorrectly, they all need to be fixed
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Well, they're not registered to fake tensor right now..
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Are you sure? I thought we hit them for dynamic shapes preferentially
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We do, you're right, but they're disabled for static fake. + 1 motivation for regularizing the path symbolic/fake (not to mention #84308). I have a pr that i had been working on to do this that i havent quite been able to push through finish line.
The aliasing tests are on for the non backward. I'll work on getting them on for backward as a follow up.
…n/var decomp aliasing (#85471) Pull Request resolved: #85471 Approved by: https://github.com/ezyang
@pytorchbot revert -m "broke tests on trunk https://hud.pytorch.org/pytorch/pytorch/commit/78afa0cf0ca04ce437ca4b519f07c04e73fe0d4c" -c nosignal |
@pytorchbot successfully started a revert job. Check the current status here. |
Reverting PR 85417 failedReason: Command
Details for Dev Infra teamRaised by workflow job |
@pytorchbot revert -m "broke tests on trunk https://hud.pytorch.org/pytorch/pytorch/commit/78afa0cf0ca04ce437ca4b519f07c04e73fe0d4c" -c nosignal |
@pytorchbot successfully started a revert job. Check the current status here. |
@eellison your PR has been successfully reverted. |
… Decomp (#85417)" This reverts commit 78afa0c. Reverted #85417 on behalf of https://github.com/clee2000 due to broke tests on trunk https://hud.pytorch.org/pytorch/pytorch/commit/78afa0cf0ca04ce437ca4b519f07c04e73fe0d4c
…ard Decomp" [ghstack-poisoned]
…ard Decomp" [ghstack-poisoned]
@pytorchbot merge |
@pytorchbot successfully started a merge job. Check the current status here. |
Hey @eellison. |
…85417) Pull Request resolved: #85417 Approved by: https://github.com/ezyang
… Decomp (#85417)" This reverts commit 9c77083. Reverted #85417 on behalf of https://github.com/clee2000 due to broke tests on trunk (and pull somehow) https://hud.pytorch.org/pytorch/pytorch/commit/9c77083965e1283763a83f72a3adf299281761e3
…85417) Pull Request resolved: #85417 Approved by: https://github.com/ezyang
… Decomp (#85417)" This reverts commit 78afa0c. Reverted #85417 on behalf of https://github.com/clee2000 due to broke tests on trunk https://hud.pytorch.org/pytorch/pytorch/commit/78afa0cf0ca04ce437ca4b519f07c04e73fe0d4c
…85417) Pull Request resolved: #85417 Approved by: https://github.com/ezyang
ghstack-source-id: 1cc1b30 Pull Request resolved: pytorch/pytorch#85417
Stack from ghstack (oldest at bottom):