Correctly set GPU device for distributed initialization. #99
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
…tibuted training. Signed-off-by: Josh Romero <joshr@nvidia.com>
Signed-off-by: Josh Romero <joshr@nvidia.com>
Signed-off-by: Josh Romero <joshr@nvidia.com>
romerojosh
force-pushed
the
fix_device_handling_distributed
branch
from
2d944f8 to
6c6554c
Compare
Collaborator
Author
|
/build_and_test |
|
🚀 Build workflow triggered! View run |
|
✅ Build workflow passed! View run |
Signed-off-by: Josh Romero <joshr@nvidia.com>
Collaborator
Author
|
/build_and_test |
|
🚀 Build workflow triggered! View run |
|
✅ Build workflow passed! View run |
azrael417
approved these changes
Dec 8, 2025
Collaborator
azrael417
left a comment
azrael417
left a comment
There was a problem hiding this comment.
Great, thanks for fixing this. I had a question but I will approve.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
The current distributed model initialization code does not set the current GPU device to the one used for the model before initializing its NCCL communicator. This can lead to potential issues where the NCCL communicator is created on a different GPU device than what the model is running on, leading to runtime crashes with NCCL operations. This PR adds appropriate device settings where required to resolve this issue.