Skip to content

Conversation

phu0ngng
Copy link
Collaborator

@phu0ngng phu0ngng commented Aug 5, 2025

Description

This PR cleans unnecessary implementations for batched_dim, sequence_dim, sequence_parallel_output, and rhs/lhs_quantized_colwise in the TE GEMM custom call.

Verified that this change does not affect E2E perf.

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

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

@phu0ngng phu0ngng marked this pull request as ready for review August 5, 2025 21:12
@phu0ngng
Copy link
Collaborator Author

phu0ngng commented Aug 5, 2025

/te-ci JAX L1

Copy link
Collaborator

@jberchtold-nvidia jberchtold-nvidia left a comment

Choose a reason for hiding this comment

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

LGTM! Thanks for this big clean up!

@phu0ngng
Copy link
Collaborator Author

phu0ngng commented Aug 6, 2025

/te-ci JAX L1

Signed-off-by: Phuong Nguyen <[email protected]>
@phu0ngng
Copy link
Collaborator Author

phu0ngng commented Aug 6, 2025

/te-ci JAX L1

@phu0ngng
Copy link
Collaborator Author

phu0ngng commented Aug 6, 2025

/te-ci JAX L1

Signed-off-by: Phuong Nguyen <[email protected]>
@phu0ngng
Copy link
Collaborator Author

phu0ngng commented Aug 7, 2025

/te-ci JAX L1

Copy link
Collaborator

@jberchtold-nvidia jberchtold-nvidia left a comment

Choose a reason for hiding this comment

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

LGTM

@phu0ngng phu0ngng merged commit cae1c43 into NVIDIA:main Aug 7, 2025
12 checks passed
@phu0ngng phu0ngng deleted the gemm_clean_up branch August 7, 2025 19:27
pggPL pushed a commit to pggPL/TransformerEngine that referenced this pull request Aug 25, 2025
* rm batch_dim, sequence_dim, sequence_parallel_output

Signed-off-by: Phuong Nguyen <[email protected]>

* rm lhs_quantized_colwise and rhs_quantized_colwise

Signed-off-by: Phuong Nguyen <[email protected]>

* rm unnecessary transpose_batch_sequence arg from some modules

Signed-off-by: Phuong Nguyen <[email protected]>

---------

Signed-off-by: Phuong Nguyen <[email protected]>
Signed-off-by: Pawel Gadzinski <[email protected]>
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