Commits · 1db649ac0204705a010ff0b2e4ec76269d3723cc · change / sglang

02 Sep, 2025 9 commits
- [feat] apply deep_gemm compile_mode to skip launch (#9879) · 1db649ac
  JieXin Liang authored Sep 02, 2025
  
  1db649ac
- fix parallel_state.py `current_platform` bug (#9919) · a1e5d781
  Xiaoyu Zhang authored Sep 02, 2025
  
  a1e5d781
- [Fix] fix the issue encountered when inference LongCat-Flash/MTP EP MoE on b200 (#9916) · b7361cc4
  Guoyuan Lin authored Sep 02, 2025
  
  b7361cc4
- Change tensor alignment method to mn major (#9844) · b9eb0d9c
  Mohammad Miadh Angkad authored Sep 02, 2025
  
  b9eb0d9c
- Move multi-tokenizer event loop to better place (#9902) · a25e8e42
  Shangming Cai authored Sep 02, 2025
```
Signed-off-by: Shangming Cai <csmthu@gmail.com>
```
  a25e8e42
- [feat] Support tp mode for DeepSeek-R1-W4AFP8 (#8118) · d4a93841
  chenxj authored Sep 02, 2025
```
Co-authored-by: yuhyao <827623970@qq.com>
```
  d4a93841
- [code style] restruct fused_moe to avoid very long single file (#9878) · b5245064
  Xiaoyu Zhang authored Sep 02, 2025
  
  b5245064
- [HiCacheStorage]: Improve 3fs kvstore‘s performance and resolve mla issues (#9876) · 58d06fdc
  hzh0425 authored Sep 02, 2025
  
  58d06fdc
- [HiCacheStorage] fix abort request host memory leaks (#9874) · cb9e0e41
  huangtingwei authored Sep 02, 2025
```
Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu>
```
  cb9e0e41
01 Sep, 2025 6 commits
- support fp8 kvcache for hybrid attn backend on GPT-OSS (#9783) · 9db80253
  Rain Jiang authored Sep 01, 2025
```
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
```
  9db80253
- Mooncake store get zero copy meta optimization (#9857) · b361750a
  huangtingwei authored Sep 01, 2025
  
  b361750a
- chore: bump v0.5.2rc0 (#9862) · 16e56ea6
  Yineng Zhang authored Sep 01, 2025
  
  16e56ea6
- chore: upgrade flashinfer 0.3.0 (#9864) · 349b491c
  Yineng Zhang authored Sep 01, 2025
  
  349b491c
- Support Multi Process Tokenizer Manager(#6555) (#8964) · 5f77e129
  ybyang authored Sep 01, 2025
```
Signed-off-by: ybyang <ybyang7@iflytek.com>
Signed-off-by: huanglong <huanglong@linux.alibaba.com>
Co-authored-by: Huang Long <121648372+LLLL114@users.noreply.github.com>
Co-authored-by: huanglong <huanglong@linux.alibaba.com>
Co-authored-by: Shangming Cai <csmthu@gmail.com>
```
  5f77e129
- Tiny allow DeepGEMM on cu12.9 (#9858) · 065e523d
  fzyzcjy authored Sep 01, 2025
  
  065e523d
31 Aug, 2025 9 commits
- fix `set_interal_state` API (#9850) · 6d3c20cf
  Liangsheng Yin authored Sep 01, 2025
  
  6d3c20cf
- [HiCache] Storage Refactoring (#9797) · 8b6966d0
  Zhiqiang Xie authored Aug 31, 2025
```
Co-authored-by: pansicheng <27603155+pansicheng@users.noreply.github.com>
```
  8b6966d0
- Fuse gate_proj and up_proj in Qwen 2.5 VL's vision MLP (#9661) · a391f73a
  Kevin Xiang Li authored Aug 31, 2025
```
Signed-off-by: Xinyuan Tong <xinyuantong.cs@gmail.com>
Co-authored-by: Xiang (Kevin) Li <lik@nvidia.com>
Co-authored-by: Xinyuan Tong <115166877+JustinTong0323@users.noreply.github.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Xinyuan Tong <xinyuantong.cs@gmail.com>
```
  a391f73a
- Fix input logprob index (#9841) · 25c73959
  Lianmin Zheng authored Aug 31, 2025
```
Co-authored-by: Sheng Shen <sheng.s@berkeley.edu>
```
  25c73959
- [HiCache] Clear kvcache in storage backend with fastAPI (#9750) · f05c6873
  Teng Ma authored Aug 31, 2025
```
Co-authored-by: hzh0425 <hzh0425@apache.org>
```
  f05c6873
- [Performance] Improve Qwen RMSNorm by replacing with native RMSNorm op (#9709) · 9a0d0b75
  Vincent Zhong authored Aug 31, 2025
  
  9a0d0b75
- [feat]Ascend NPU Gemma-3-12b and Gemma-3-27b support (#8909) · ba861293
  VDV1985 authored Aug 31, 2025
  
  ba861293
- [Model] Support Meituan LongCat-Flash && LongCat-Flash-MTP (#9824) · 5e194b21
  Guoyuan Lin authored Aug 31, 2025
  
  5e194b21
- hot fix for mooncake batch set api (#9836) · f9076a5a
  Zhiqiang Xie authored Aug 30, 2025
  
  f9076a5a
30 Aug, 2025 11 commits
- [Auto Sync] Update parallel_state.py (20250830) (#9828) · 1e61b496
  Lianmin Zheng authored Aug 30, 2025
```
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
```
  1e61b496
- chore: upgrade transformers 4.56.0 (#9827) · 300676af
  Yineng Zhang authored Aug 30, 2025
  
  300676af
- chore: upgrade sgl-kernel 0.3.7.post1 with deepgemm fix (#9822) · 9970e3bf
  Yineng Zhang authored Aug 30, 2025
  
  9970e3bf
- Fix typo in warning message about DeepGEMM JIT (#9802) · 70eedb58
  Mohammad Miadh Angkad authored Aug 30, 2025
  
  70eedb58
- Fix memory leak when aborting decode request in PD-Disagg (#9817) · 836873b9
  Liangsheng Yin authored Aug 30, 2025
```
Co-authored-by: Lianmin Zheng <15100009+merrymercy@users.noreply.github.com>
```
  836873b9
- fix: dsv3 lite q_lora_rank none (#9815) · 8abe8dea
  Yineng Zhang authored Aug 29, 2025
  
  8abe8dea
- feature(eplb): add min-rebalancing-utilization-threshold for eplb (#8345) · c2a26e72
  hzh0425 authored Aug 30, 2025
```
Co-authored-by: yizhang2077 <1109276519@qq.com>
```
  c2a26e72
- Small bug fix in transformers model implementation (#9809) · 591e6c59
  yilian49 authored Aug 29, 2025
  
  591e6c59
- Adds initialize_moe_config to bench_one_batch so MOE backend is respected (#9670) · 42f34437
  pranavm-nvidia authored Aug 29, 2025
  
  42f34437
- Fix TRTLLM MLA Cuda KV Blocks Causing accuracy drop (#9675) · ff9b5618
  Faraz authored Aug 29, 2025
  
  ff9b5618
- [ModelOpt] Fix Weight Loading for DSR1-FP4 Quantization (#9712) · fcd72bd1
  Pavani Majety authored Aug 29, 2025
```
Signed-off-by: Pavani Majety <pmajety@nvidia.com>
```
  fcd72bd1
29 Aug, 2025 5 commits

chore: upgrade flashinfer 0.3.0rc1 (#9793) · 3d8fc434
Yineng Zhang authored Aug 29, 2025

3d8fc434
[feat] Support EAGLE3 for Qwen2 (#9216) · 87a0f7d2
KerwinKai authored Aug 30, 2025

87a0f7d2

feat: add original logprobs to response (#8375) · 839c93bd

narutolhy authored Aug 29, 2025


Co-authored-by: Chayenne <zhaochen20@outlook.com>
Co-authored-by: luhongyu.4869 <luhongyu.4869@bytedance.com>

839c93bd

feat: Add flexible validation for partial weight updates (#9663) · f1e9bbaf

JiLi authored Aug 30, 2025


Co-authored-by: RichardW <rich-junwang@users.noreply.github.com>
Co-authored-by: Zhuorany <yzr1914001753@gmail.com>
Co-authored-by: Stefan He <hebiaobuaa@gmail.com>
Co-authored-by: Yineng Zhang <me@zhyncs.com>
Co-authored-by: Night <32424487+PrinsYin@users.noreply.github.com>
Co-authored-by: zhaochenyang20 <zhaochen20@outlook.com>
Co-authored-by: Liangsheng Yin <hnyls2002@gmail.com>

f1e9bbaf

support enable in the reasoning field to enable thingking for thinkin… (#9715) · 3fd1431d
gongwei-130 authored Aug 29, 2025

3fd1431d