Commits · 229d2b95f19573ece9c1c5d6b357df9874e04f59 · zhaoyu6 / sglang

30 Sep, 2025 3 commits
- [CPU] Adding Memory Capacity Acquisition Functionality (#11102) · 229d2b95
  Zaili Wang authored Sep 30, 2025
  
  229d2b95
- Enable optional FP32 compute for LM Head (#10729) · d17986f8
  narutolhy authored Sep 29, 2025
```
Thanks to MiniMax Team and Chenyang Zhao's support.
```
  d17986f8
- 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 4 commits
- Fix mem fraction static for nightly tests (#11076) · dda34c2f
  Lianmin Zheng authored Sep 29, 2025
  
  dda34c2f
- Use more general heuristics to set the default value of --mem-fraction-static (#10975) · a17e70f5
  Lianmin Zheng authored Sep 29, 2025
```
Co-authored-by: sglang-bot <sglangbot@gmail.com>
```
  a17e70f5
- [speculative decoding] rename lookahead to ngram (#11010) · 24f7cb1e
  Zhihao Zhang authored Sep 29, 2025
```
Co-authored-by: a4zhangfei <a4zhangfei@qq.com>
```
  24f7cb1e
- [HiCacheStorage] mooncake store support page_first_direct layout (#10591) · e05555fa
  huangtingwei authored Sep 29, 2025
  
  e05555fa
28 Sep, 2025 1 commit
- fix: show failed models in nightly ci (#10986) · 2e763398
  Mick authored Sep 29, 2025
  
  2e763398
27 Sep, 2025 2 commits

fix: make inference deterministic for large TP (#10930) · 62e2e99d

Xinyuan Tong authored Sep 27, 2025


Co-authored-by: yhyang201 <yhyang201@gmail.com>
Co-authored-by: Yangmin Li <yangminl@nvidia.com>
Co-authored-by: Yuan Luo <yuan.luo@hotmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

62e2e99d

[HiCache]: Support dynamic loading backends for hicache (#10551) · c8a5d12a
hzh0425 authored Sep 27, 2025
```
Co-authored-by: Teng Ma <sima.mt@alibaba-inc.com>
```
c8a5d12a

26 Sep, 2025 2 commits
- Restruct gpu_memory_settings in a unify function and relax max_cuda_graph_bs (#10372) · 05a35266
  Xiaoyu Zhang authored Sep 27, 2025
```
Co-authored-by: Lianmin Zheng <lianminzheng@gmail.com>
Co-authored-by: sglang-bot <sglangbot@gmail.com>
```
  05a35266
- [PD-HiCache]: Support Async Offloading KVCache In Decode Side (#10192) · 7ec5b4e8
  hzh0425 authored Sep 26, 2025
```
Signed-off-by: Shangming Cai <csmthu@gmail.com>
Co-authored-by: Shangming Cai <csmthu@gmail.com>
```
  7ec5b4e8
25 Sep, 2025 2 commits
- Rename customer label -> custom label (#10899) · f68dd998
  Lianmin Zheng authored Sep 25, 2025
```
Co-authored-by: Yingchun Lai <laiyingchun@apache.org>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
```
  f68dd998
- integrate AIBrix KVcache (#10376) · fce17048
  yi wang authored Sep 25, 2025
  
  fce17048
24 Sep, 2025 1 commit

chore: Initial support for input config files (#10534) · d7b20dd6

kushanam authored Sep 24, 2025


Co-authored-by: root <root@umbriel-b200-017.ipp4a1.colossus.nvidia.com>
Co-authored-by: Yineng Zhang <me@zhyncs.com>

d7b20dd6

23 Sep, 2025 2 commits
- Clean up server args (#10770) · ea338676
  Lianmin Zheng authored Sep 23, 2025
  
  ea338676
- [Ascend] optimize Qwen-vl on Ascend (#10556) · 095093ee
  ronnie_zheng authored Sep 23, 2025
```
Co-authored-by: wangqihui01 <wangqh10@163.com>
```
  095093ee
22 Sep, 2025 3 commits
- Add warnings and remove dependency for deterministic inference (#10724) · aa1c5cf5
  Baizhou Zhang authored Sep 22, 2025
```
Co-authored-by: Yineng Zhang <me@zhyncs.com>
```
  aa1c5cf5
- [2/2] Support deterministic inference for temperature > 0 (#10678) · e2ac7888
  Qiaolin Yu authored Sep 21, 2025
```
Co-authored-by: Baizhou Zhang <sobereddiezhang@gmail.com>
Co-authored-by: hebiao064 <hebiaobuaa@gmail.com>
```
  e2ac7888
- Support deterministic inference with triton backend (#10694) · 134b4f7e
  Ethan (Yusheng) Su authored Sep 21, 2025
  
  134b4f7e
21 Sep, 2025 4 commits
- [3/4] Speed up CSGMV backend perf by 10% through dynamic chunking + kernel optimization (#10592) · 08ecd0aa
  Lifu Huang authored Sep 20, 2025
  
  08ecd0aa
- Super tiny fix extra logs (#10697) · 720c1c8c
  fzyzcjy authored Sep 21, 2025
  
  720c1c8c
- feat: update server args (#10696) · d403c143
  Yineng Zhang authored Sep 20, 2025
```
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
```
  d403c143
- [Feature] Support deterministic inference with FA3 backend (#10651) · cba0d8c3
  Stefan He authored Sep 20, 2025
  
  cba0d8c3
20 Sep, 2025 1 commit
- [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
18 Sep, 2025 3 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
- 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
- [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
17 Sep, 2025 2 commits
- 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
- feat: add priority based scheduling with priority based request acceptance and preemption (#8746) · 14fdd527
  harrisonlimh authored Sep 16, 2025
  
  14fdd527
16 Sep, 2025 2 commits
- support using fa4 on deepseek on blackwell (#9928) · a2f7218a
  cicirori authored Sep 17, 2025
  
  a2f7218a
- [CPU] fix CPU backend sel. issue for Llama4 (#10511) · 925dbb32
  Zaili Wang authored Sep 16, 2025
  
  925dbb32
15 Sep, 2025 5 commits
- [2/2] Introduce Chunked-SGMV kernels and corresponding LoRA backend for... · 3f41b48c
  Lifu Huang authored Sep 15, 2025
```
[2/2] Introduce Chunked-SGMV kernels and corresponding LoRA backend for improved performance (#10286)
```
  3f41b48c
- fix(server_args): Skip chunked_prefill_size validation when disaggregation mode is decode (#10358) · 3795b6a4
  Jimmy authored Sep 15, 2025
  
  3795b6a4
- metrics: support customer labels specified in request header (#10143) · fc2c3a3d
  Yingchun Lai authored Sep 15, 2025
  
  fc2c3a3d
- [4/N]DP refactor: support watching mode `get_load` and shortest queue strategy (#10201) · 305c9e8c
  Liangsheng Yin authored Sep 15, 2025
  
  305c9e8c
- Add split tile size for Triton attention (#10425) · 60d7beda
  Ke Bao authored Sep 15, 2025
  
  60d7beda
14 Sep, 2025 2 commits
- [Feature] Sglang Tracing: Fine-Grained Tracking for Request Latency - Part 1 (#9962) · 4c21b090
  Feng Su authored Sep 15, 2025
```
Signed-off-by: Feng Su <sufeng@linux.alibaba.com>
Signed-off-by: Huaixin Chang <changhuaixin@linux.alibaba.com>
Signed-off-by: Peng Wang <rocking@linux.alibaba.com>
```
  4c21b090
- [Hack] Add pd-disaggregation decode polling interval (#10411) · 55a6e644
  Liangsheng Yin authored Sep 14, 2025
  
  55a6e644
13 Sep, 2025 1 commit
- [Performance] Dynamic Batch Tokenizer (#9382) · 94d0f656
  Sundara Raman Ramachandran authored Sep 13, 2025
  
  94d0f656