[functorch] excise older custom_vjp prototype #84638
Conversation
It was based off of the Python op registration API that has been implemented in PyTorch already, so we can always bring it back, but we're probably taking a different approach here. Test Plan: - tests [ghstack-poisoned]
🔗 Helpful links
✅ No Failures (20 Pending)As of commit cbdac93 (more details on the Dr. CI page): Expand to see more💚 💚 Looks good so far! There are no failures yet. 💚 💚 This comment was automatically generated by Dr. CI (expand for details).Please report bugs/suggestions to the (internal) Dr. CI Users group. |
It was based off of the Python op registration API that has been implemented in PyTorch already, so we can always bring it back, but we're probably taking a different approach here. Test Plan: - tests [ghstack-poisoned]
It was based off of the Python op registration API that has been implemented in PyTorch already, so we can always bring it back, but we're probably taking a different approach here. Test Plan: - tests [ghstack-poisoned]
It was based off of the Python op registration API that has been implemented in PyTorch already, so we can always bring it back, but we're probably taking a different approach here. Test Plan: - tests [ghstack-poisoned]
|
@zou3519 has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
| from functorch.experimental import functionalize | ||
|
|
||
| if not IS_WINDOWS: | ||
| from functorch._src.custom_function import custom_vjp |
There was a problem hiding this comment.
Do we want to get rid of the file "functorch/_src/custom_function.py" too?
It was based off of the Python op registration API that has been implemented in PyTorch already, so we can always bring it back, but we're probably taking a different approach here. Test Plan: - tests Differential Revision: [D39315050](https://our.internmc.facebook.com/intern/diff/D39315050) [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/84638
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 03aa4d3: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
@zou3519 has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
zou3519
added
the
test-config/functorch
It was based off of the Python op registration API that has been implemented in PyTorch already, so we can always bring it back, but we're probably taking a different approach here. Test Plan: - tests Differential Revision: [D39315050](https://our.internmc.facebook.com/intern/diff/D39315050) [ghstack-poisoned]
zou3519
removed
the
test-config/functorch
It was based off of the Python op registration API that has been implemented in PyTorch already, so we can always bring it back, but we're probably taking a different approach here. Test Plan: - tests Differential Revision: [D39315050](https://our.internmc.facebook.com/intern/diff/D39315050) [ghstack-poisoned]
Summary: X-link: pytorch/pytorch#84638 It was based off of the Python op registration API that has been implemented in PyTorch already, so we can always bring it back, but we're probably taking a different approach here. ghstack-source-id: 166807991 Reviewed By: samdow Differential Revision: D39315050 Pulled By: zou3519 fbshipit-source-id: 9831cdea0e6401d89137811104f463094b6bca3f
|
@pytorchbot merge (Initiating merge automatically since Phabricator Diff has merged) |
|
@pytorchbot successfully started a merge job. Check the current status here. |
|
Hey @zou3519. |
Summary: Pull Request resolved: #84638 It was based off of the Python op registration API that has been implemented in PyTorch already, so we can always bring it back, but we're probably taking a different approach here. ghstack-source-id: 166807991 Test Plan: - tests Reviewed By: samdow Differential Revision: D39315050 Pulled By: zou3519 fbshipit-source-id: 9831cdea0e6401d89137811104f463094b6bca3f
4 participants
Stack from ghstack (oldest at bottom):
It was based off of the Python op registration API that has been
implemented in PyTorch already, so we can always bring it back, but
we're probably taking a different approach here.
Test Plan:
Differential Revision: D39315050