Fix losses and aggregators when using CUDA graphs#280
Merged
ktangsali merged 10 commits intoNVIDIA:mainfrom Feb 19, 2026
Merged
Fix losses and aggregators when using CUDA graphs#280ktangsali merged 10 commits intoNVIDIA:mainfrom
ktangsali merged 10 commits intoNVIDIA:mainfrom
Conversation
Signed-off-by: Jason Ye <jasonyecanada@gmail.com>
Signed-off-by: Jason Ye <jasonyecanada@gmail.com>
Signed-off-by: Jason Ye <jasonyecanada@gmail.com>
Merge in main repo
ktangsali
reviewed
Feb 13, 2026
Collaborator
ktangsali
left a comment
ktangsali
left a comment
There was a problem hiding this comment.
Hi @jasooney23 , thank you for your PR. I have started to take a look at this and we will keep you posted soon
loliverhennigh
approved these changes
Feb 18, 2026
Collaborator
loliverhennigh
left a comment
loliverhennigh
left a comment
There was a problem hiding this comment.
I went throught the larger code blocks fairly carefully and seems correct to me. I would say if the unit tests run and we were able to do some sanity check on the examples we are good to go
Collaborator
|
/blossom-ci |
Refactor forward method signature for better readability.
Collaborator
|
/blossom-ci |
ktangsali
approved these changes
Feb 19, 2026
Collaborator
ktangsali
left a comment
ktangsali
left a comment
There was a problem hiding this comment.
We ran a few end-to-end tests too to verify, and everything works. CI also passes, so I don't have any issues merging this fix.
Thank you for addressing the issue
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
PhysicsNeMo Pull Request
Description
closes #279
Passes
stepas an int scalar tensor to gradient computation functions so that CUDA graphs can track it properly.Refactors loss functions and aggregators to replace
ifstatements withtorch.wheres for CUDA graph compatibility.Checklist
Dependencies