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
Tests for verifying behaviour of BatchNorm using 0-dim batch sizes. #32384
Conversation
💊 CircleCI build failures summary and remediationsAs of commit d123d0e: Commit d123d0e was recently pushed. Waiting for builds... This comment was automatically generated by Dr. CI (expand for details).Follow this link to opt-out of these comments for your Pull Requests.Please report bugs/suggestions on the GitHub issue tracker. This comment has been revised 22 times. |
There is already work in #30035. However that PR does not check that the gradients of weight & bias should be zero rather than None; it does not check moving_mean & moving_var; it does not check the gradient w.r.t input. It might be better to improve upon the existing one with extra checks. |
I've updated the previous test. However the exact behaviour of the |
@ppwwyyxx can you please review this? |
WIP for now. Will ping once I'm sure the tests are passing. |
@rgommers can you please have a look at this one? |
@ppwwyyxx this one seems ready to merge. Do have a look. |
This looks good to me but I don't have permission to merge pytorch diffs. |
LGTM too. |
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.
@ngimel has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
…ytorch#32384) Summary: The `BatchNorm*` part of the issue (see pytorchgh-12013) seems to have been fixed in the master branch and these tests would make it concrete. However I would appreciate comments on pytorch#12013 (comment) on whether the current behaviour is satisfactory. Pull Request resolved: pytorch#32384 Differential Revision: D19704154 Pulled By: ngimel fbshipit-source-id: 1bbbbf1ae1215a460b22cf26e6b263e518ecf60b
…ytorch#32384) Summary: The `BatchNorm*` part of the issue (see pytorchgh-12013) seems to have been fixed in the master branch and these tests would make it concrete. However I would appreciate comments on pytorch#12013 (comment) on whether the current behaviour is satisfactory. Pull Request resolved: pytorch#32384 Differential Revision: D19704154 Pulled By: ngimel fbshipit-source-id: 1bbbbf1ae1215a460b22cf26e6b263e518ecf60b
The
BatchNorm*
part of the issue (see gh-12013) seems to have been fixed in the master branch and these tests would make it concrete.However I would appreciate comments on #12013 (comment) on whether the current behaviour is satisfactory.