Commits · 106c2b31fb8a1a7ebc0ea3c1447a80ae03ef37d3 · change / sglang · GitLab

04 Sep, 2025 11 commits
- feat(hicache): Add generic hicache ci e2e test and benchmark test (#9846) · 106c2b31
  hzh0425 authored Sep 04, 2025
```
Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu>
```
  106c2b31
- Ensure chunked request extension length respects both rem_chunk_tokens and... · c6756949
  pansicheng authored Sep 04, 2025
```
Ensure chunked request extension length respects both rem_chunk_tokens and rem_total_tokens limits (#10003)
```
  c6756949
- [1/N] DP-refactor: move dp balance code into scheduler's mixin class (#10004) · 27e8ffed
  Liangsheng Yin authored Sep 04, 2025
  
  27e8ffed
- fix: health_generate endpoint in mini_lb (#9997) · 4dbb34fe
  wxsm authored Sep 04, 2025
  
  4dbb34fe
- [Bugfix] fix pd chat completion protocol for batching support (#10016) · 1e18a341
  Tony Lu authored Sep 04, 2025
```
Signed-off-by: Tony Lu <tonylu@linux.alibaba.com>
```
  1e18a341
- Fix Llama 4 with MXFP4 dynamic quant on MI35x (#9993) · 2c562fd2
  Hubert Lu authored Sep 04, 2025
  
  2c562fd2
- [Fix] gpt-oss mxfp4 model run failed on ROCm platform (#9994) · b648d862
  kk authored Sep 04, 2025
```
Co-authored-by: wunhuang <wunhuang@amd.com>
```
  b648d862
- [router] fix grpc connection mode detection (#9999) · bbf261ae
  Simo Lin authored Sep 04, 2025
  
  bbf261ae
- [router] clean up dependency injector to use ctx (#10000) · 4f8a982d
  Simo Lin authored Sep 04, 2025
  
  4f8a982d
- [router] move tokenizer, reasoning, tool initialization to server (#9996) · d966b902
  Simo Lin authored Sep 03, 2025
  
  d966b902
- feat: add gpt oss b200 ci (#9988) · de921733
  Yineng Zhang authored Sep 03, 2025
  
  de921733
03 Sep, 2025 21 commits
- [Auto Sync] Update parallel_state.py, few_shot_gsm8k.py (20250903) (#9986) · 397448eb
  Lianmin Zheng authored Sep 03, 2025
```
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Leon Gao <leon.gao19@gmail.com>
```
  397448eb
- Minor update regarding issue #9704 (#9733) · 66d5d042
  Elfie Guo authored Sep 03, 2025
  
  66d5d042
- nsys profile output kernel classifier (#9314) · 73179b76
  Grace Ho authored Sep 03, 2025
```
Signed-off-by: Grace Ho <grho@nvidia.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Yineng Zhang <me@zhyncs.com>
```
  73179b76
- Triton 3.4.0 MoE config for Deepseek TP16 H100 (#9978) · 8cbf71dc
  Szymon Ożóg authored Sep 03, 2025
  
  8cbf71dc
- fix swa clear(): rename is_in_free_group to is_not_in_free_group (#9914) · 56eb5d0a
  Xinyuan Tong authored Sep 03, 2025
  
  56eb5d0a
- Remove mrope position sync (#9460) · 4ed9053e
  timmy-feng authored Sep 03, 2025
```
Co-authored-by: Nathan Wang <nathan.r.wang@gmail.com>
```
  4ed9053e
- [router] add chat_template_kwargs in ChatCompletionRequest (#9958) · 5e19b159
  Tony Lu authored Sep 04, 2025
```
Signed-off-by: Tony Lu <tonylu@linux.alibaba.com>
```
  5e19b159
- [router] Add Rerank API Specification (#9906) · 788b19a5
  Frank Fang authored Sep 03, 2025
  
  788b19a5
- [1/N][Bug] Fix w4afp8 MoE NaN issue (sgl-kernel) (#9953) · f78b7fd1
  Yuhao Yao authored Sep 03, 2025
  
  f78b7fd1
- [benchmark] add flashinfer_allreduce_fusion benchmark (#9937) · b1fb7e45
  Xiaoyu Zhang authored Sep 03, 2025
  
  b1fb7e45
- Revert "Optimized deepseek-v3/r1 model performance on mxfp4 run (#9671)" (#9959) · 1b2ff4fb
  Yineng Zhang authored Sep 03, 2025
  
  1b2ff4fb
- Revert "[Fix] DeepSeek EP accuracy issue on B200 GPUs (#9946)" (#9955) · 2c7ca33a
  Yineng Zhang authored Sep 02, 2025
  
  2c7ca33a
- [feat] Add P/D attention select for draft model (#9755) · df397a72
  Ximingwang-09 authored Sep 03, 2025
```
Co-authored-by: 纬杭 <ximing.wxm@antgroup.com>
```
  df397a72
- add proctitle for tokenizers (#9952) · 5dfcd6c2
  Liangsheng Yin authored Sep 03, 2025
```
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
```
  5dfcd6c2
- Optimized deepseek-v3/r1 model performance on mxfp4 run (#9671) · 0dfd54d1
  kk authored Sep 03, 2025
```
Co-authored-by: wunhuang <wunhuang@amd.com>
Co-authored-by: wghuang <wghuang@amd.com>
```
  0dfd54d1
- Qwen FP8/NVFP4 ModelOPT Quantization support (#7912) · bcbeed71
  jingyu-ml authored Sep 02, 2025
```
Co-authored-by: Jingyu Xin <jingyux@nvidia.com>
```
  bcbeed71
- Update tool_chat_template_deepseekv31.jinja (#9895) · cc9a31c6
  南京小汤包 authored Sep 03, 2025
  
  cc9a31c6
- Remove annoying warnings in sgl kernel build (#9905) · d631290e
  Lianmin Zheng authored Sep 02, 2025
  
  d631290e
- fix(cache): move ongoing_prefetch pop after validation to prevent leak (#9927) · 37565b7f
  JinYan Su authored Sep 03, 2025
```
Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu>
```
  37565b7f
- [Fix] DeepSeek EP accuracy issue on B200 GPUs (#9946) · 6243c367
  Al-Ekram Elahee Hridoy authored Sep 02, 2025
  
  6243c367
- Move parsers under a single folder (#9912) · 60e37f80
  Lianmin Zheng authored Sep 02, 2025
  
  60e37f80
02 Sep, 2025 8 commits
- [HiCache] Minor fix on file storage backend (#9869) · 369b1433
  Zhiqiang Xie authored Sep 02, 2025
  
  369b1433
- [model] support MiniCPM-V 4.0 (#8747) · 03dbf1aa
  tc-mb authored Sep 03, 2025
```
Signed-off-by: tc-mb <caitianchi@modelbest.cn>
Co-authored-by: Xinyuan Tong <115166877+JustinTong0323@users.noreply.github.com>
```
  03dbf1aa
- Grpc client (#9939) · 11dcabc5
  Chang Su authored Sep 02, 2025
  
  11dcabc5
- Fix the key passing issue in page first layout. (#9929) · 4d89389c
  hzh0425 authored Sep 03, 2025
  
  4d89389c
- [router] include rust benchamrks (#9932) · 9491d6e5
  Simo Lin authored Sep 02, 2025
  
  9491d6e5
- Support the internvl3.5 family models in sglang (#9705) · f64b8e3e
  yilian49 authored Sep 02, 2025
  
  f64b8e3e
- [Hicache] Generic page get bugfix (#9909) · 53976fce
  ykwd authored Sep 02, 2025
  
  53976fce
- chore: bump v0.5.2rc1 (#9920) · 18f91eb6
  Yineng Zhang authored Sep 02, 2025
  
  18f91eb6