Skip to content

[JAX-Q] Distributed MXFP8 flax layer tests#1643

Merged
phu0ngng merged 2 commits intoNVIDIA:mainfrom
jberchtold-nvidia:dev/jberchtold/mxfp8-layer-test
Apr 4, 2025
Merged

[JAX-Q] Distributed MXFP8 flax layer tests#1643
phu0ngng merged 2 commits intoNVIDIA:mainfrom
jberchtold-nvidia:dev/jberchtold/mxfp8-layer-test

Conversation

@jberchtold-nvidia
Copy link
Collaborator

Description

Fixes an issue with the Flax layer axes so distributed MXFP8 layers work correctly and the distributed layer tests can be enabled.

Type of change

  • Documentation change (change only to the documentation, either a fix or a new content)
  • Bug fix (non-breaking change which fixes an issue)
  • New feature (non-breaking change which adds functionality)
  • Breaking change (fix or feature that would cause existing functionality to not work as expected)
  • Infra/Build change
  • Code refactoring

Changes

Please list the changes introduced in this PR:

  • Update axis partitioning in transformer_engine/jax/flax/module.py
  • Enable distributed layer tests in tests/jax/test_distributed_layernorm_mlp.py

Checklist:

  • I have read and followed the contributing guidelines
  • The functionality is complete
  • I have commented my code, particularly in hard-to-understand areas
  • I have made corresponding changes to the documentation
  • My changes generate no new warnings
  • I have added tests that prove my fix is effective or that my feature works
  • New and existing unit tests pass locally with my changes

@jberchtold-nvidia jberchtold-nvidia force-pushed the dev/jberchtold/mxfp8-layer-test branch from 44e0d23 to a9b5e95 Compare April 3, 2025 19:18
@jberchtold-nvidia
Copy link
Collaborator Author

/te-ci L1

Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>
@jberchtold-nvidia jberchtold-nvidia force-pushed the dev/jberchtold/mxfp8-layer-test branch from a9b5e95 to 603b452 Compare April 3, 2025 19:29
@jberchtold-nvidia
Copy link
Collaborator Author

/te-ci L1

@phu0ngng phu0ngng self-requested a review April 3, 2025 22:38
Copy link
Collaborator

@phu0ngng phu0ngng left a comment

Choose a reason for hiding this comment

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

LGTM!
Congrats on your "first" PR!

@phu0ngng phu0ngng merged commit be1f647 into NVIDIA:main Apr 4, 2025
11 checks passed
lhb8125 pushed a commit to lhb8125/TransformerEngine that referenced this pull request Apr 8, 2025
MXFP8 flax layer tests

Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>
wdykas pushed a commit to wdykas/TransformerEngine that referenced this pull request Apr 14, 2025
MXFP8 flax layer tests

Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>
Signed-off-by: Peter Dykas <wdykas@nvidia.com>
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