Skip to content

Nsys JAX update for remapping nccl nodes#1829

Merged
Steboss merged 3 commits intomainfrom
sbosisio/fix_nsys_jax
Dec 17, 2025
Merged

Nsys JAX update for remapping nccl nodes#1829
Steboss merged 3 commits intomainfrom
sbosisio/fix_nsys_jax

Conversation

@Steboss
Copy link
Contributor

@Steboss Steboss commented Dec 9, 2025

Updated data_loaders.py so we can:

  • find all the NCCL range IDs
  • delete the NCCL rows
  • remap the IDs correctly to preserve the correct ids for the logic XlaModule --> Thunk: Operation

@Steboss Steboss requested a review from olupton December 9, 2025 13:14
@olupton
Copy link
Collaborator

olupton commented Dec 10, 2025

There seem to be a bunch of new test failures.

@Steboss
Copy link
Contributor Author

Steboss commented Dec 17, 2025

@olupton
it looks like the latest PRs solved the issue here

Copy link
Collaborator

@olupton olupton left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks! I think it should be OK not to update the RangeStack values at this point.

@Steboss Steboss merged commit 6dd41d1 into main Dec 17, 2025
60 of 65 checks passed
@Steboss Steboss deleted the sbosisio/fix_nsys_jax branch December 17, 2025 14:39
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants