Fixed handling of nan for evenly_distribute_backward #45280
Conversation
[ghstack-poisoned]
heitorschueroff
changed the title
heitorschueroff
added
the
module: autograd
albanD
left a comment
albanD
left a comment
There was a problem hiding this comment.
Looks good!
Maybe add a test to make sure it has the right behavior now both on CPU and CUDA?
I'll add some test now but more will come on the PR for the nan ops. |
Performance is the same on CPU and on CUDA is only 1-1.05x slower. This change is necessary for the future nan ops including nan(min|max|median) [ghstack-poisoned]
💊 CI failures summary and remediationsAs of commit 95b1638 (more details on the Dr. CI page): 💚 💚 Looks good so far! There are no failures yet. 💚 💚 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 or post in the (internal) Dr. CI Users group. This comment has been revised 10 times. |
Performance is the same on CPU and on CUDA is only 1-1.05x slower. This change is necessary for the future nan ops including nan(min|max|median) [ghstack-poisoned]
Performance is the same on CPU and on CUDA is only 1-1.05x slower. This change is necessary for the future nan ops including nan(min|max|median) [ghstack-poisoned]
Performance is the same on CPU and on CUDA is only 1-1.05x slower. This change is necessary for the future nan ops including nan(min|max|median) Differential Revision: [D23908796](https://our.internmc.facebook.com/intern/diff/D23908796) [ghstack-poisoned]
Codecov Report
@@ Coverage Diff @@
## gh/heitorschueroff/16/base #45280 +/- ##
===========================================================
Coverage 68.05% 68.05%
===========================================================
Files 396 396
Lines 51232 51232
===========================================================
Hits 34865 34865
Misses 16367 16367 Continue to review full report at Codecov.
|
|
@heitorschueroff merged this pull request in 96f8755. |
Stack from ghstack:
Performance is the same on CPU and on CUDA is only 1-1.05x slower. This change is necessary for the future nan ops including nan(min|max|median)
Differential Revision: D23908796