Commits · 81fd2b0ee0df11f5d7db486d7bf1180f4f5bd196 · change / sglang

22 Oct, 2025 1 commit
- [sgl-kernel] support flashmla libtorch (#11717) · 23afdfd1
  Fan Yin authored Oct 22, 2025
  
  23afdfd1
21 Oct, 2025 1 commit
- [NVIDIA] Add new SMs support for Spark & Thor (#11287) · 2b1da821
  Serge Panev authored Oct 21, 2025
```
Signed-off-by: Serge Panev <spanev@nvidia.com>
```
  2b1da821
20 Oct, 2025 2 commits
- [DeepseekV32] Add fast_topk_transform_ragged_fused kernel (#11815) · 3b80232d
  hlu1 authored Oct 19, 2025
```
Signed-off-by: Hao Lu <14827759+hlu1@users.noreply.github.com>
```
  3b80232d
- [NVIDIA] FA3/FA4 Fix (#11606) · 252dc4e1
  Johnny authored Oct 20, 2025
```
Co-authored-by: Baizhou Zhang <sobereddiezhang@gmail.com>
```
  252dc4e1
16 Oct, 2025 1 commit
- [sgl-kernel] support hadamard (#11663) · 3289da5b
  Fan Yin authored Oct 16, 2025
  
  3289da5b
15 Oct, 2025 1 commit
- [sgl-kernel][3/N]Support Expert Specialization Grouped GEMM (#11674) · 6c01844f
  Qi Yuhang authored Oct 16, 2025
  
  6c01844f
14 Oct, 2025 1 commit
- Revert "[NVIDIA] BUMP FA3 (#11444)" (#11582) · f792e3c5
  Yineng Zhang authored Oct 13, 2025
  
  f792e3c5
13 Oct, 2025 3 commits

[sgl-kernel][2/N]Support Expert Specialization Grouped GEMM (#11534) · dc48c4c0
Qi Yuhang authored Oct 14, 2025

dc48c4c0

Johnny authored Oct 13, 2025


Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: ishandhanani <82981111+ishandhanani@users.noreply.github.com>

b8c430f1

[sgl-kernel][1/N]Support Expert Specialization Grouped GEMM (#11432) · 9a30914e

Qi Yuhang authored Oct 13, 2025


Co-authored-by: luoyuan.luo <luoyuan.luo@antgroup.com>
Co-authored-by: PGFLMG <1106310035@qq.com>
Co-authored-by: Xiaoyu Zhang <35585791+BBuf@users.noreply.github.com>

9a30914e

11 Oct, 2025 2 commits
- [7/n] decouple quantization impl from vllm dependency - gguf kernel (#11019) · 8fdcd98e
  PGFLMG authored Oct 12, 2025
  
  8fdcd98e
- Reland [1/2] Optimizations and refactors about quant kernel (#10312) · 21337b22
  fzyzcjy authored Oct 11, 2025
```
Co-authored-by: Yineng Zhang <me@zhyncs.com>
```
  21337b22
07 Oct, 2025 1 commit
- [sgl-kernel] Support float64 moe_sum_reduce cuda kernel (#11068) · 4f42c8cd
  Yuan Luo authored Oct 07, 2025
```
Co-authored-by: luoyuan.luo <luoyuan.luo@antgroup.com>
```
  4f42c8cd
05 Oct, 2025 1 commit
- [Feature] Add a fast-topk to sgl-kernel for DeepSeek v3.2 (#11194) · e0b2d3ee
  DarkSharpness authored Oct 06, 2025
```
Co-authored-by: Baizhou Zhang <sobereddiezhang@gmail.com>
```
  e0b2d3ee
29 Sep, 2025 1 commit
- [speculative decoding] rename lookahead to ngram (#11010) · 24f7cb1e
  Zhihao Zhang authored Sep 29, 2025
```
Co-authored-by: a4zhangfei <a4zhangfei@qq.com>
```
  24f7cb1e
28 Sep, 2025 1 commit

[sgl-kernel] Optimize concat_mla_k kernel (#10543) · 42245551

Yuan Luo authored Sep 28, 2025


Co-authored-by: luoyuan.luo <luoyuan.luo@antgroup.com>
Co-authored-by: PGFLMG <1106310035@qq.com>

42245551

25 Sep, 2025 1 commit
- [Bug] Fix Issue#10215 (#10572) · fe531d6f
  Yuhao Yao authored Sep 25, 2025
  
  fe531d6f
22 Sep, 2025 1 commit
- Optimize cutlass int8 gemm kernel for large M on SM89 Ada GPU (#10714) · 0f04a5f4
  Qi Yuhang authored Sep 22, 2025
  
  0f04a5f4
19 Sep, 2025 1 commit

[sgl-kernel] Support moe_sum_reduce cuda kernel (#10321) · 616a3e20

Yuan Luo authored Sep 19, 2025


Co-authored-by: luoyuan.luo <luoyuan.luo@antgroup.com>
Co-authored-by: Xiaoyu Zhang <35585791+BBuf@users.noreply.github.com>

616a3e20

18 Sep, 2025 1 commit
- [Feature] Speculative decoding support lookahead (#9873) · e7bc6003
  Zhihao Zhang authored Sep 19, 2025
```
Co-authored-by: a4zhangfei <a4zhangfei@qq.com>
Co-authored-by: Qiaolin-Yu <liin1211@outlook.com>
```
  e7bc6003
16 Sep, 2025 1 commit
- Update CUTLASS. Refine KernelSchedule for fp8 (grouped) gemm. (#10491) · 9b876889
  Qi Yuhang authored Sep 16, 2025
  
  9b876889
15 Sep, 2025 2 commits
- [1/2] Speed up trtllm_mla attention backend (>10% e2e) (#10473) · 3b25dc12
  fzyzcjy authored Sep 16, 2025
  
  3b25dc12
- Fix correction bias undefined behavior for nvfp4 models (#10426) · 258d02c8
  fzyzcjy authored Sep 15, 2025
  
  258d02c8
13 Sep, 2025 1 commit
- Fix the style of sgl kernel (#10398) · c9ec4cae
  Lianmin Zheng authored Sep 12, 2025
  
  c9ec4cae
11 Sep, 2025 2 commits
- Fix errors of hicache kernels in sgl-kernel for ROCm (#10339) · fe68c148
  Hubert Lu authored Sep 11, 2025
  
  fe68c148
- Revert "[1/2] Optimizations and refactors about quant kernel (#9534)" (#10292) · 6d55f60e
  Yineng Zhang authored Sep 10, 2025
  
  6d55f60e
10 Sep, 2025 2 commits
- pass a_scale from fp8 quant result instead of hard code to 1.0f (#10241) · 2286e85e
  Rain Jiang authored Sep 10, 2025
```
Co-authored-by: Yichen Wang <yichen.wang@bytedance.com>
Co-authored-by: Jinwu Guo <641876696@qq.com>
```
  2286e85e
- Page first direct IO kernel (#10060) · 5be8c2f7
  huangtingwei authored Sep 10, 2025
```
Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu>
```
  5be8c2f7
09 Sep, 2025 2 commits
- Add mamba kernel (#10234) · 8cbe1538
  Yi Zhang authored Sep 10, 2025
  
  8cbe1538
- [CPU] Add gelu_and_mul kernel in sgl-kernel and add ut (#9300) · d1d4074c
  blzheng authored Sep 09, 2025
  
  d1d4074c
08 Sep, 2025 3 commits
- [1/2] Speed up prefill mla attention (#10156) · 0096798e
  fzyzcjy authored Sep 09, 2025
  
  0096798e
- [1/N][Bug] Fix w4afp8 MoE NaN issue (sgl-kernel, fixed) (#10108) · ee0b3c5b
  Yuhao Yao authored Sep 08, 2025
  
  ee0b3c5b
- Add graph runner support with torch compile on CPU (#7843) · 7577f0e4
  Cao E authored Sep 08, 2025
  
  7577f0e4
07 Sep, 2025 2 commits
- Optimize nvfp4 block scaled gemm kernel when M is small. (#10101) · 85ed8e0a
  Qi Yuhang authored Sep 07, 2025
  
  85ed8e0a
- CUTLASS fp8 blockwise gemm support of sm120 (#9969) · dd1e2689
  Jianying authored Sep 07, 2025
  
  dd1e2689
06 Sep, 2025 4 commits
- [chore] Remove unused ep_moe cuda kernels (#9956) · 5f1eb204
  hlu1 authored Sep 06, 2025
  
  5f1eb204
- Disable kernel cutlass_mla_decode on SM103 (#10058) · 4c22ebe2
  hlu1 authored Sep 06, 2025
```
Signed-off-by: Hao Lu <14827759+hlu1@users.noreply.github.com>
```
  4c22ebe2
- [Fix] illegal sync based on undefined behaviour (#9620) · dbb1235d
  DevashishLal-CB authored Sep 05, 2025
```
Signed-off-by: Devashish Lal <devashish@rivosinc.com>
Co-authored-by: Xiaoyu Zhang <35585791+BBuf@users.noreply.github.com>
```
  dbb1235d
- Revert "[1/N][Bug] Fix w4afp8 MoE NaN issue (sgl-kernel) (#9953)" (#10097) · 0e78c63c
  Yineng Zhang authored Sep 05, 2025
  
  0e78c63c
05 Sep, 2025 1 commit
- Support copying tensor from cpu to gpu without using copy engines (#10007) · bd7f8821
  fzyzcjy authored Sep 05, 2025
  
  bd7f8821