Commits · 546445725118b57e1365cbf34578f4a7ca4de018 · change / sglang

15 Oct, 2025 2 commits
- [sgl-kernel] Optimize gguf test (#11667) · 54644572
  Fan Yin authored Oct 16, 2025
  
  54644572
- [sgl-kernel][3/N]Support Expert Specialization Grouped GEMM (#11674) · 6c01844f
  Qi Yuhang authored Oct 16, 2025
  
  6c01844f
13 Oct, 2025 1 commit

[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
06 Oct, 2025 1 commit
- [Bug] Fix incorrect assertion in FA4 and add UT. (#11182) · 748f86f3
  Lifu Huang authored Oct 06, 2025
  
  748f86f3
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
25 Sep, 2025 1 commit
- [Bug] Fix Issue#10215 (#10572) · fe531d6f
  Yuhao Yao authored Sep 25, 2025
  
  fe531d6f
24 Sep, 2025 1 commit
- Restruct sgl-kernel benchmark (#10861) · c4e314f9
  Xiaoyu Zhang authored Sep 25, 2025
  
  c4e314f9
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
11 Sep, 2025 1 commit
- 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 3 commits
- Add mamba kernel (#10234) · 8cbe1538
  Yi Zhang authored Sep 10, 2025
  
  8cbe1538
- feat: support fa cute in sgl-kernel (#10205) · 94fb4e9e
  Yineng Zhang authored Sep 09, 2025
```
Co-authored-by: cicirori <32845984+cicirori@users.noreply.github.com>
```
  94fb4e9e
- feat: use sgl-kernel cu129 as default (#10188) · cdc56ef6
  Yineng Zhang authored Sep 08, 2025
  
  cdc56ef6
08 Sep, 2025 1 commit
- [1/N][Bug] Fix w4afp8 MoE NaN issue (sgl-kernel, fixed) (#10108) · ee0b3c5b
  Yuhao Yao authored Sep 08, 2025
  
  ee0b3c5b
06 Sep, 2025 2 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
05 Sep, 2025 1 commit
- [1/2] Optimizations and refactors about quant kernel (#9534) · 339f8eef
  fzyzcjy authored Sep 05, 2025
  
  339f8eef
02 Sep, 2025 1 commit
- [feat] Support tp mode for DeepSeek-R1-W4AFP8 (#8118) · d4a93841
  chenxj authored Sep 02, 2025
```
Co-authored-by: yuhyao <827623970@qq.com>
```
  d4a93841
30 Aug, 2025 1 commit
- [NVIDIA] [2/N] Optimize `silu_and_mul_scaled_fp4_grouped_quant` perf (#9556) · 5c34b4f1
  Kaixi Hou authored Aug 29, 2025
  
  5c34b4f1
29 Aug, 2025 1 commit
- Skip some tests on Blackwell (#9777) · a7d825fc
  hlu1 authored Aug 28, 2025
```
Signed-off-by: Hao Lu <14827759+hlu1@users.noreply.github.com>
```
  a7d825fc
25 Aug, 2025 1 commit
- Update CUTLASS 4.2 & Enable K-Major Scale Factor for SM90 FP8 Blockwise Group GEMM (#9559) · fda47926
  Qi Yuhang authored Aug 25, 2025
  
  fda47926
22 Aug, 2025 1 commit
- [NVIDA] [1/N] Nvfp4 Masked Gemm: Add quant op for the flashinfer grouped gemm (#9200) · e5638573
  Kaixi Hou authored Aug 22, 2025
  
  e5638573
21 Aug, 2025 1 commit
- Fix quant kernel test errors and benchmark wrong output speeds (#7604) · e85cb1ce
  fzyzcjy authored Aug 21, 2025
  
  e85cb1ce
20 Aug, 2025 2 commits
- [2/2] Fuse routed scaling factor into select_experts (#8690) · a91e90d9
  Trevor Morris authored Aug 20, 2025
  
  a91e90d9
- Reduce overhead for fa by not calling heavy CUDA property check (#7375) · c9bf3877
  Yichen Yan authored Aug 20, 2025
  
  c9bf3877
14 Aug, 2025 2 commits
- [sgl-kernel] Support FlashInfer top_k_top_p_sampling_from_logits (#9060) · 53dcc750
  Yuan Luo authored Aug 15, 2025
```
Co-authored-by: luoyuan.luo <luoyuan.luo@antgroup.com>
```
  53dcc750
- [2/n]decouple quantization implementation from vLLM dependency (#8112) · 5aa1ebd2
  Peng Zhang authored Aug 14, 2025
```
Co-authored-by: walker-ai <yiyun.wyt@antgroup.com>
Co-authored-by: leoneo <1320612015@qq.com>
```
  5aa1ebd2
13 Aug, 2025 1 commit
- Update fa3 interface and add unit test (#9150) · 94f44b88
  Ke Bao authored Aug 13, 2025
  
  94f44b88
12 Aug, 2025 1 commit
- Fuse writing KV buffer into rope kernel (part 1: sgl-kernel) (#9077) · 9aea2555
  fzyzcjy authored Aug 12, 2025
  
  9aea2555
11 Aug, 2025 1 commit
- Revert "[1/2][resubmit] sgl-kernel: Fuse routed scaling factor into m… (#9035) · dd949ace
  Yineng Zhang authored Aug 10, 2025
  
  dd949ace
09 Aug, 2025 1 commit
- [1/2][resubmit] sgl-kernel: Fuse routed scaling factor into moe_fused_gate (select_experts) (#8770) · 591c232f
  Trevor Morris authored Aug 08, 2025
  
  591c232f
08 Aug, 2025 1 commit
- refactor: Move scalar_types.py to sgl-kernel to avoid circular import (#8720) · 39fd1788
  Hongbo Xu authored Aug 08, 2025
  
  39fd1788
02 Aug, 2025 2 commits
- Revert "[1/2] sgl-kernel: Fuse routed scaling factor into select_experts" (#8706) · f9f0138f
  Liangsheng Yin authored Aug 02, 2025
  
  f9f0138f
- [1/2] sgl-kernel: Fuse routed scaling factor into select_experts (#8364) · f642524f
  Trevor Morris authored Aug 01, 2025
  
  f642524f
01 Aug, 2025 2 commits
- fix per token cuda kernel hidden dim cannot divide by 16 (#8543) · db7343c9
  Stefan He authored Aug 01, 2025
  
  db7343c9
- [Kimi K2] dsv3_router_gemm supports NUM_EXPERTS == 384 (#8013) · 6bdd2786
  Peter Pan authored Aug 01, 2025
  
  6bdd2786