Commits · 2a2ff9a8407f78c3af085fe3d4f76aab0ba58dae · change / sglang

"examples/speech_recognition/utils/wer_utils.py" did not exist on "33646ac9b7b720444416e2c18d1120d03b37e156"

18 Sep, 2025 19 commits
- refactor: use registry for _get_attention_backend_from_str (#10629) · 2a2ff9a8
  Yineng Zhang authored Sep 18, 2025
  
  2a2ff9a8
- [router] refactor worker to builder pattern 1/n (#10628) · 5291f32d
  Simo Lin authored Sep 18, 2025
  
  5291f32d
- Garbage collector regression in the online server (#10621) · 67073dde
  brayden-hai authored Sep 18, 2025
  
  67073dde
- 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
- Fix bias handling in TritonMoeQuantInfo within quantization/mxfp4.py (#10579) · 388c05d5
  yhyang201 authored Sep 19, 2025
  
  388c05d5
- [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
- Fix sgl_kernel import failure on devices other than CUDA (#10610) · 6fd4816d
  Zaili Wang authored Sep 19, 2025
  
  6fd4816d
- support qwen3-next-fp8 deepep (#10622) · 1344ebc8
  Yi Zhang authored Sep 19, 2025
  
  1344ebc8
- update deepep version for qwen3-next deepep moe (#10624) · e07b21ce
  Yi Zhang authored Sep 19, 2025
  
  e07b21ce
- Feat/add heartbeat mechanism for nixl conn (#10222) · 52f248cd
  shaharmor98 authored Sep 18, 2025
```
Signed-off-by: Shahar Mor <smor@nvidia.com>
```
  52f248cd
- [RL] Add destroy process group api (#9979) · 93f75778
  penguin_wwy authored Sep 19, 2025
  
  93f75778
- fix deepep assert when PD disaggregation == null (#8274) · 4039c626
  AlphaBaby authored Sep 19, 2025
```
Co-authored-by: fujianhao.fjh <fujianhao.fjh@antgroup.com>
```
  4039c626
- Scale kkt after reduction (#10604) · db71c38f
  Yi Zhang authored Sep 18, 2025
  
  db71c38f
- [improvement] add average input/output token length for hicache benchmark stats output (#10525) · 7a68b422
  zhannngchen authored Sep 18, 2025
  
  7a68b422
- chore: upgrade mooncake 0.3.6 (#10596) · 60fc5b51
  Shangming Cai authored Sep 18, 2025
```
Signed-off-by: Shangming Cai <csmthu@gmail.com>
```
  60fc5b51
- [PD] Improve disaggregation common backend and refactor mooncake backend (#10273) · a13dd1e4
  Shangming Cai authored Sep 18, 2025
```
Signed-off-by: Shangming Cai <csmthu@gmail.com>
```
  a13dd1e4
- aiter v0.1.5.post2 (#10563) · d500eb91
  HAI authored Sep 17, 2025
  
  d500eb91
- [HICache] introduce evict policy (#10190) · 1ccd59c7
  Xuchun Shang authored Sep 18, 2025
```
Signed-off-by: Xuchun Shang <xuchun.shang@linux.alibaba.com>
Co-authored-by: Teng Ma <sima.mt@alibaba-inc.com>
```
  1ccd59c7
- [ROCm] Fix fp8 quantization accuracy issue. (#10558) · c32fb7a2
  sogalin authored Sep 18, 2025
  
  c32fb7a2
17 Sep, 2025 14 commits
- Enable trtllm mla prefix extend (#10526) · 1ba137e9
  Shu Wang authored Sep 17, 2025
  
  1ba137e9
- vlm: remove redundant d2h movement of mm feature tensors (#9987) · de28f8e7
  Kevin Xiang Li authored Sep 17, 2025
```
Co-authored-by: Xiang (Kevin) Li <lik@nvidia.com>
```
  de28f8e7
- fix: update dsv3 fp4 ut (#10584) · 56405076
  Yineng Zhang authored Sep 17, 2025
  
  56405076
- [BugFix] Fix incorrect hidden_states_tensor in pd disaggregation + eagle (#9976) · b73ac629
  Ziming Huang authored Sep 18, 2025
  
  b73ac629
- [HiCache] Add tests for hicache storage mooncake backend (#10171) · 77098aea
  Teng Ma authored Sep 18, 2025
```
Signed-off-by: Shangming Cai <csmthu@gmail.com>
Co-authored-by: hzh0425 <hzh0425@apache.org>
Co-authored-by: Shangming Cai <csmthu@gmail.com>
```
  77098aea
- [bench] Fix random seed in `bench_one_batch_server` (#10548) · 5ccf0b03
  Liangsheng Yin authored Sep 17, 2025
  
  5ccf0b03
- CUDA Arch Independent (#8813) · a77564e0
  EduardDurech authored Sep 17, 2025
  
  a77564e0
- Remove duplicated code (#10545) · 4f9e71df
  Yichen Yan authored Sep 17, 2025
```
Co-authored-by: jiapingW <56055330+jiapingW@users.noreply.github.com>
```
  4f9e71df
- [bugfix]hicache bench_long_context.py run failed (#10523) · 541551ce
  zhannngchen authored Sep 17, 2025
  
  541551ce
- enable prefix cache with dp (#10459) · 124097fc
  Shu Wang authored Sep 16, 2025
  
  124097fc
- Fix decord dependency for aarch64 docker build (#10529) · e1d45bc2
  kyleliang-nv authored Sep 16, 2025
  
  e1d45bc2
- feat: add priority based scheduling with priority based request acceptance and preemption (#8746) · 14fdd527
  harrisonlimh authored Sep 16, 2025
  
  14fdd527
- [Auto Sync] Update activation.py, chunk_cache.py, utils.py (20250917) (#10538) · f949ad57
  Lianmin Zheng authored Sep 16, 2025
```
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
```
  f949ad57
- [Auto Sync] Update scheduler_profiler_mixin.py, rpd_utils.p... (20250916) (#10494) · c49484a6
  Lianmin Zheng authored Sep 16, 2025
```
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: cctry <shiyang@x.ai>
```
  c49484a6
16 Sep, 2025 7 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
- skip vision_model for lora (#10530) · 373080ea
  gongwei-130 authored Sep 16, 2025
  
  373080ea
- Fix formatting in long code blocks (#10528) · 7f028b07
  Philip Kiely - Baseten authored Sep 16, 2025
  
  7f028b07
- adjust import setuptools_rust (#10524) · 0abb41c7
  ybyang authored Sep 16, 2025
  
  0abb41c7
- [CPU] fix CPU backend sel. issue for Llama4 (#10511) · 925dbb32
  Zaili Wang authored Sep 16, 2025
  
  925dbb32
- Support sgl-router parallel_batch in bench_one_batch_server (#10506) · 8df7353a
  fzyzcjy authored Sep 16, 2025
  
  8df7353a