OpInfo for adaptive_avg_pool2d#62704
Conversation
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit 3ed1f38 (more details on the Dr. CI page): ✅ None of the CI failures appear to be your fault 💚
❄️ 1 failure tentatively classified as flakybut reruns have not yet been triggered to confirm:
|
zou3519
left a comment
There was a problem hiding this comment.
This looks pretty good, I had one small suggestion for a test case
| ((1, 8, 8, 8), (5, 7)), | ||
| ((2, 8, 8, 8), (None, 7)), | ||
| ((1, 8, 4, 3), (5, None)), | ||
| ((1, 8, 4, 3), (None, None)), |
There was a problem hiding this comment.
output_size is allowed to be a single integer according to https://pytorch.org/docs/stable/generated/torch.nn.functional.adaptive_avg_pool2d.html?highlight=adaptive_avg_pool#torch.nn.functional.adaptive_avg_pool2d, could we test that case for completeness?
There was a problem hiding this comment.
Thanks for taking a look, @zou3519 - appreciate the suggestion. I've added a case in the latest commit, and have also merged with upstream. :)
|
@zou3519 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
|
This pull request has been reverted by 6e24ce7. |
Summary: This PR is an attempt to reland #62704. **What has changed?** The op has non-deterministic behavior, hence an appropriate `gradcheck` wrapper had to be added. cc: mruberry zou3519 heitorschueroff kshitij12345 Pull Request resolved: #62935 Reviewed By: anjali411 Differential Revision: D30225095 Pulled By: zou3519 fbshipit-source-id: 644873cc21d44b19c8b68f9edff691913778de0e
Summary: This PR is an attempt to reland #62704. **What has changed?** The op has non-deterministic behavior, hence an appropriate `gradcheck` wrapper had to be added. cc: mruberry zou3519 heitorschueroff kshitij12345 Pull Request resolved: #62935 Reviewed By: anjali411 Differential Revision: D30225095 Pulled By: zou3519 fbshipit-source-id: 644873cc21d44b19c8b68f9edff691913778de0e
Please see pytorch/functorch#78 and #54261.
Note regarding sample inputs for this function:
output_size:(None, None), (None, width), (height, None), (height, width).cc: @mruberry @zou3519 @Chillee