Skip to content

[Pytorch] MOE Sequential Block - #3350

Open
vthumbe1503 wants to merge 138 commits into
NVIDIA:mainfrom
vthumbe1503:dispatch_combine_op
Open

[Pytorch] MOE Sequential Block#3350
vthumbe1503 wants to merge 138 commits into
NVIDIA:mainfrom
vthumbe1503:dispatch_combine_op

Conversation

@vthumbe1503

Copy link
Copy Markdown
Collaborator

Description

Please include a brief summary of the changes, relevant motivation and context.

Fixes # (issue)

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:

  • Change A
  • Change B

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

vthumbe1503 and others added 30 commits July 28, 2026 23:05
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
…h error handling tests

Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
…/TransformerEngine into enable_extra_out_consumption
@ptrendx ptrendx added the 2.20 label Aug 31, 2026
vthumbe1503 and others added 5 commits August 31, 2026 15:38
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Signed-off-by: root <root@hecate0161.hecate.clusters.nvidia.com>
const int64_t *const first_dims_ptr = reinterpret_cast<const int64_t *>(input->first_dims.dptr);
const int64_t *const last_dims_ptr = reinterpret_cast<const int64_t *>(input->last_dims.dptr);

// Prefer the rowwise representation when both GEMM orientations are available.

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

P2 Dual-Orientation Path Untested

This newly supported path selects the rowwise data and scales when both grouped MXFP8 orientations are present, but the dequantization tests cover only each orientation separately and still state that both is invalid. Without a both case that verifies the intended rowwise selection, this behavior can regress without detection.

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

Signed-off-by: vthumbe <vthumbe@nvidia.com>
Signed-off-by: vthumbe <vthumbe@nvidia.com>
Signed-off-by: Varun Thumbe <vthumbe@nvidia.com>
Comment thread transformer_engine/pytorch/ops/fused/moe_ep.py Outdated
Signed-off-by: vthumbe <vthumbe@nvidia.com>
Comment thread transformer_engine/pytorch/ops/fused/moe_ep.py Outdated
Signed-off-by: vthumbe <vthumbe@nvidia.com>
Signed-off-by: vthumbe <vthumbe@nvidia.com>
Signed-off-by: vthumbe <vthumbe@nvidia.com>
Signed-off-by: vthumbe <vthumbe@nvidia.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants