Commits · 20a6c0a63d643825e4be5c809fa685490a25ec0d · zhaoyu6 / sglang

"vscode:/vscode.git/clone" did not exist on "6b1897d6752f3926a91d0af07deb84258030b26d"

12 Oct, 2025 1 commit

Beta spec-overlap for EAGLE (#11398) · 20a6c0a6

Liangsheng Yin authored Oct 12, 2025


Co-authored-by: Lianmin Zheng <15100009+merrymercy@users.noreply.github.com>
Co-authored-by: Hanming Lu <69857889+hanming-lu@users.noreply.github.com>

20a6c0a6

11 Oct, 2025 1 commit
- [DPSKv3.2] Rewrite nsa tilelang act_quant kernel to triton (#11450) · 451d15c4
  Binyao Jiang authored Oct 10, 2025
  
  451d15c4
09 Oct, 2025 3 commits
- move more files under srt/utils (#11285) · 9b8ebb27
  Lianmin Zheng authored Oct 09, 2025
  
  9b8ebb27
- chore: upgrade flashinfer 0.4.0 (#11364) · 44cb0607
  Yineng Zhang authored Oct 09, 2025
  
  44cb0607
- [Generative Score API] Multi-Item scoring with custom attention mask. (#10979) · 53bd00d9
  Sundara Raman Ramachandran authored Oct 08, 2025
  
  53bd00d9
08 Oct, 2025 2 commits
- model: Support Hybrid Mamba2 NemotronHForCausalLM (nvidia/NVIDIA-Nemotron-Nano-9B-v2) (#10909) · d6837aea
  Netanel Haber authored Oct 08, 2025
```
Signed-off-by: Netanel Haber <nhaber@nvidia.com>
```
  d6837aea
- [2/2] Support MHA prefill with FlashAttention 4. (#10937) · edefab0c
  Lifu Huang authored Oct 08, 2025
```
Co-authored-by: Hieu Pham <hyhieu@gmail.com>
```
  edefab0c
06 Oct, 2025 1 commit

Support DeepSeek V3.2 Exp (#11061) · efbc687c

fzyzcjy authored Oct 06, 2025


Co-authored-by: Stefan He <11166516+hebiao064@users.noreply.github.com>
Co-authored-by: Liangsheng Yin <95566987+hnyls2002@users.noreply.github.com>
Co-authored-by: Baizhou Zhang <56809903+fridge003@users.noreply.github.com>
Co-authored-by: DarkSharpness <76582120+darksharpness@users.noreply.github.com>
Co-authored-by: ZhengdQin <46387172+zhengdqin@users.noreply.github.com>
Co-authored-by: DarkSharpness <2040703891@qq.com>
Co-authored-by: hnyls2002 <lsyincs@gmail.com>
Co-authored-by: Zhengda Qin <zhengdqin@gmail.com>
Co-authored-by: Liangsheng Yin <hnyls2002@gmail.com>
Co-authored-by: HAI <hixiao@gmail.com>
Co-authored-by: Baizhou Zhang <sobereddiezhang@gmail.com>

efbc687c

04 Oct, 2025 1 commit
- [fix]enable flashmla when using draft model P/D attention select (#11012) · 666da3d5
  Hank Han authored Oct 04, 2025
  
  666da3d5
03 Oct, 2025 1 commit
- [1/n] Enable DCA CUDA graph capture (#9537) · a2faf894
  b8zhong authored Oct 02, 2025
  
  a2faf894
02 Oct, 2025 3 commits
- Allow use of TRTLLM_MHA backend for hybrid attention on Blackwell (#11138) · e8100774
  Dom Brown authored Oct 03, 2025
  
  e8100774
- [model] Add mamba2 and Falcon-H1 support. (#10988) · 083629c2
  ilyasch2 authored Oct 02, 2025
```
Co-authored-by: Younes Belkada <younes.belkada@tii.ae>
Co-authored-by: Younes B <49240599+younesbelkada@users.noreply.github.com>
```
  083629c2
- Fuse quantize and rope in trtllm_mla MTP (#10779) · f35def86
  fzyzcjy authored Oct 02, 2025
  
  f35def86
01 Oct, 2025 2 commits
- feat: add fast_decode_plan from flashinfer, flashinfer to 0.4.0rc3 (#10760) · ac1f2928
  eigen authored Oct 01, 2025
```
Co-authored-by: Zihao Ye <yezihhhao@gmail.com>
Co-authored-by: Sleepcoo <Sleepcoo@gmail.com>
```
  ac1f2928
- Organize spec-related data structures (#10735) · 73d4a5f8
  Liangsheng Yin authored Oct 01, 2025
  
  73d4a5f8
30 Sep, 2025 2 commits
- [Fix] Resolve performance drop in speculative decoding aiter backend (#11087) · 59911195
  jacky.cheng authored Sep 30, 2025
  
  59911195
- Remove hybrid_linear_attn attention backend and refactor attention registry (#10816) · 2bc61dd1
  li-kesen authored Sep 30, 2025
```
Co-authored-by: Yi Zhang <1109276519@qq.com>
```
  2bc61dd1
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
23 Sep, 2025 3 commits
- Fix trtllm_mla slow concat kernel in MTP (#10777) · 063c3791
  fzyzcjy authored Sep 23, 2025
  
  063c3791
- [Ascend] optimize Qwen-vl on Ascend (#10556) · 095093ee
  ronnie_zheng authored Sep 23, 2025
```
Co-authored-by: wangqihui01 <wangqh10@163.com>
```
  095093ee
- [Feature] Add MLAProcess for DeepSeek MLA on NPU (#10130) · d27a6f70
  Even Zhou authored Sep 23, 2025
  
  d27a6f70
22 Sep, 2025 3 commits
- [Auto Sync] Update flashattention_backend.py (20250922) (#10762) · 0753ef83
  Yineng Zhang authored Sep 22, 2025
```
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Gordon Gustafson <ggustafson@together.ai>
```
  0753ef83
- Enables TRT-LLM backend to be used for target_verify (#10281) · b1bb8e74
  pranavm-nvidia authored Sep 22, 2025
```
Co-authored-by: Pranav Marathe <pranavm@ipp1-3309.ipp1a1.colossus.nvidia.com>
Co-authored-by: fzyzcjy <ch271828n@outlook.com>
```
  b1bb8e74
- Support deterministic inference with triton backend (#10694) · 134b4f7e
  Ethan (Yusheng) Su authored Sep 21, 2025
  
  134b4f7e
21 Sep, 2025 1 commit
- [Feature] Support deterministic inference with FA3 backend (#10651) · cba0d8c3
  Stefan He authored Sep 20, 2025
  
  cba0d8c3
20 Sep, 2025 2 commits
- [Performance] Qwen3-Next: optimize causal_conv1d_fn triton kernel - up to 9% faster (#10680) · bd7eb020
  Binyao Jiang authored Sep 20, 2025
  
  bd7eb020
- [1/2] Support deterministic inference with flashinfer attention backend (#10645) · 8ecef73f
  Baizhou Zhang authored Sep 19, 2025
```
Co-authored-by: hebiao064 <hebiaobuaa@gmail.com>
Co-authored-by: Qiaolin-Yu <liin1211@outlook.com>
```
  8ecef73f
19 Sep, 2025 4 commits
- limit sgl-kernel causal conv1d to cuda only (#10648) · d88ef4a3
  Jinyan Chen authored Sep 20, 2025
```
Co-authored-by: Jinyan Chen <jinyanc@nvidia.com>
```
  d88ef4a3
- Fix fast decode plan for flashinfer v0.4.0rc1 and upgrade sgl-kernel 0.3.11 (#10634) · 3fa3c22a
  Baizhou Zhang authored Sep 19, 2025
```
Co-authored-by: zhyncs <me@zhyncs.com>
```
  3fa3c22a
- [Performance] Qwen3-Next: speed up update_mamba_state_after_mtp_verify by 10x;... · 4e6c4923
  Binyao Jiang authored Sep 18, 2025
```
[Performance] Qwen3-Next: speed up update_mamba_state_after_mtp_verify by 10x; e2e up to 3.54% faster (#10586)
```
  4e6c4923
- [Performance] Qwen3-Next: replace arange to cached query_start_loc_li… (#10553) · b91cb67e
  Binyao Jiang authored Sep 18, 2025
  
  b91cb67e
18 Sep, 2025 5 commits
- [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
- refactor: use registry for _get_attention_backend_from_str (#10629) · 2a2ff9a8
  Yineng Zhang authored Sep 18, 2025
  
  2a2ff9a8
- feat: Add FlexAttention Backend for Efficient Sparse Attention (#9947) · 9a5c42f9
  yuk.igalaxy authored Sep 19, 2025
```
Co-authored-by: Baizhou Zhang <sobereddiezhang@gmail.com>
```
  9a5c42f9
- [Performance] qwen3-next improve causal conv1d in prefill phase (#10595) · fc809665
  Jinyan Chen authored Sep 19, 2025
```
Co-authored-by: Jinyan Chen <jinyanc@nvidia.com>
```
  fc809665
- Scale kkt after reduction (#10604) · db71c38f
  Yi Zhang authored Sep 18, 2025
  
  db71c38f
17 Sep, 2025 1 commit
- Enable trtllm mla prefix extend (#10526) · 1ba137e9
  Shu Wang authored Sep 17, 2025
  
  1ba137e9
16 Sep, 2025 3 commits
- support using fa4 on deepseek on blackwell (#9928) · a2f7218a
  cicirori authored Sep 17, 2025
  
  a2f7218a
- [2/2] Speed up trtllm_mla attention backend (#10474) · 311de47b
  fzyzcjy authored Sep 17, 2025
  
  311de47b
- feat: update support for qwen3next model (#10466) · b6dd4bcb
  cao1zhg authored Sep 16, 2025
  
  b6dd4bcb