[PR #369] [MERGED] support indexer_n_heads = 16 #394

Closed
opened 2026-10-26 21:17:13 +00:00 by bsakaguchi · 0 comments
Owner

📋 Pull Request Information

Original PR: https://github.com/deepseek-ai/DeepGEMM/pull/369
Author: @JaredforReal
Created: 6/26/2026
Status: ✅ Merged
Merged: 6/29/2026
Merged by: @RayWang96

Base: nv_dev ← Head: feat/indexer_n_heads_16


📝 Commits (1)

  • ac7324d support indexer_n_heads = 16

📊 Changes

6 files changed (+36 additions, -28 deletions)

View changed files

📝 csrc/apis/attention.hpp (+4 -4)
📝 csrc/jit_kernels/impls/smxx_fp8_fp4_paged_mqa_logits.hpp (+4 -2)
📝 deep_gemm/include/deep_gemm/impls/sm100_fp4_mqa_logits.cuh (+5 -6)
📝 deep_gemm/include/deep_gemm/impls/sm100_fp4_paged_mqa_logits.cuh (+9 -7)
📝 deep_gemm/include/deep_gemm/impls/sm100_fp8_mqa_logits.cuh (+5 -4)
📝 deep_gemm/include/deep_gemm/impls/sm100_fp8_paged_mqa_logits.cuh (+9 -5)

📄 Description

migrated from https://github.com/vllm-project/DeepGEMM/pull/1


🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.

## 📋 Pull Request Information **Original PR:** https://github.com/deepseek-ai/DeepGEMM/pull/369 **Author:** [@JaredforReal](https://github.com/JaredforReal) **Created:** 6/26/2026 **Status:** ✅ Merged **Merged:** 6/29/2026 **Merged by:** [@RayWang96](https://github.com/RayWang96) **Base:** `nv_dev` ← **Head:** `feat/indexer_n_heads_16` --- ### 📝 Commits (1) - [`ac7324d`](https://github.com/deepseek-ai/DeepGEMM/commit/ac7324d0ad5a352d2017fdd589c25921ce667d60) support indexer_n_heads = 16 ### 📊 Changes **6 files changed** (+36 additions, -28 deletions) <details> <summary>View changed files</summary> 📝 `csrc/apis/attention.hpp` (+4 -4) 📝 `csrc/jit_kernels/impls/smxx_fp8_fp4_paged_mqa_logits.hpp` (+4 -2) 📝 `deep_gemm/include/deep_gemm/impls/sm100_fp4_mqa_logits.cuh` (+5 -6) 📝 `deep_gemm/include/deep_gemm/impls/sm100_fp4_paged_mqa_logits.cuh` (+9 -7) 📝 `deep_gemm/include/deep_gemm/impls/sm100_fp8_mqa_logits.cuh` (+5 -4) 📝 `deep_gemm/include/deep_gemm/impls/sm100_fp8_paged_mqa_logits.cuh` (+9 -5) </details> ### 📄 Description migrated from https://github.com/vllm-project/DeepGEMM/pull/1 --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>
bsakaguchi 2026-10-26 21:17:13 +00:00
Sign in to join this conversation.
No labels
pull-request
No milestone
No project
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#394
No description provided.