deepseek-ai/DeepGEMM
1
0
Fork 0
mirror of https://github.com/deepseek-ai/DeepGEMM.git synced 2026-11-05 00:47:26 +00:00
Code Issues 162 Projects Packages Wiki Activity

[GH-ISSUE #104] Grouped WGRAD kernel #58

New issue
Closed
opened 2026-10-26 21:11:29 +00:00 by bsakaguchi · 2 comments
bsakaguchi commented 2026-10-26 21:11:29 +00:00
Owner
Copy link

Originally created by @cassanof on GitHub (May 20, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepGEMM/issues/104

Hey! I took a look at the new WGRAD kernel; really nice release! from what it looks like, there isn't a grouped kernel yet, are you planning to release one in the future?

Thanks!

Originally created by @cassanof on GitHub (May 20, 2025). Original GitHub issue: https://github.com/deepseek-ai/DeepGEMM/issues/104 Hey! I took a look at the new WGRAD kernel; really nice release! from what it looks like, there isn't a grouped kernel yet, are you planning to release one in the future? Thanks!
bsakaguchi commented 2026-10-26 21:11:31 +00:00
Author
Owner
Copy link

@zheanxu commented on GitHub (May 21, 2025):

The k grouped weight gradient GEMM kernel has already been released. Please check:
github.com/deepseek-ai/DeepGEMM@391755ada0/deep_gemm/jit_kernels/wgrad_gemm.py (L116)

<!-- gh-comment-id:2896271166 --> @zheanxu commented on GitHub (May 21, 2025): The k grouped weight gradient GEMM kernel has already been released. Please check: https://github.com/deepseek-ai/DeepGEMM/blob/391755ada0ffefa9a6a52b6f14dcaf22d1a463e0/deep_gemm/jit_kernels/wgrad_gemm.py#L116
bsakaguchi commented 2026-10-26 21:11:31 +00:00
Author
Owner
Copy link

@cassanof commented on GitHub (May 21, 2025):

Hey, I am a bit confused on how it's grouped, since it's just looping for every expert?

<!-- gh-comment-id:2896449417 --> @cassanof commented on GitHub (May 21, 2025): Hey, I am a bit confused on how it's grouped, since it's just looping for every expert?
Sign in to join this conversation.
No Branch/Tag specified
Branches Tags
main
public-sync-20261014
0930-release
nv_dev
fix-mega-moe-fence
public-release-2609
public-release-2607
public-release-2606
mega-update
public-release-260416
ko3n1g/ci/build-wheels
nv_dev_f8e8fb5
nv_dev_950d31a
nv_dev_67fc648
nv_dev_c491439
nv_dev_4ff3f54
v2.1.1.post3
v2.1.1.post2
v2.1.1.post1
v2.1.1
v2.1.0
Labels
Clear labels   
pull-request
Mirrored from GitHub Pull Request

No labels
pull-request
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
deepseek-ai/DeepGEMM#58
No description provided.