enable cudagraph for AUCMetricComputation #2951
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
facebook-github-bot
added
the
CLA Signed
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D74293458 |
BoyuanFeng
force-pushed
the
export-D74293458
branch
from
5049f40 to
6731994
Compare
BoyuanFeng
added a commit
to BoyuanFeng/torchrec
that referenced
this pull request
May 15, 2025
Summary: Previously, enable cudagraph leads to [errors](https://fb.workplace.com/groups/1075192433118967/permalink/1661777797793758/) on AUCMetricComputation. The root cause is that model forward output tensors `predictions`, `labels`, `weights`, and `grouping_keys` got overwritten by the next CUDAGraph replay. This diff fixes the issue by clone these tensors before they are overwritten. The overhead should be small since these tensors only have shape `(n_task, n_examples)`. Reviewed By: TroyGarden Differential Revision: D74293458
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D74293458 |
Summary: Pull Request resolved: meta-pytorch#2951 Previously, enable cudagraph leads to [errors](https://fb.workplace.com/groups/1075192433118967/permalink/1661777797793758/) on AUCMetricComputation. The root cause is that model forward output tensors `predictions`, `labels`, `weights`, and `grouping_keys` got overwritten by the next CUDAGraph replay. This diff fixes the issue by clone these tensors before they are overwritten. The overhead should be small since these tensors only have shape `(n_task, n_examples)`. Reviewed By: iamzainhuda, TroyGarden Differential Revision: D74293458
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D74293458 |
BoyuanFeng
force-pushed
the
export-D74293458
branch
from
6731994 to
79064a2
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary:
Previously, enable cudagraph leads to errors on AUCMetricComputation.
The root cause is that model forward output tensors
predictions,labels,weights, andgrouping_keysgot overwritten by the next CUDAGraph replay. This diff fixes the issue by clone these tensors before they are overwritten. The overhead should be small since these tensors only have shape
(n_task, n_examples).Differential Revision: D74293458