Commits · ac1f2928aef9d5af3cd86f413a9e2626a3bc2f1d · change / sglang · GitLab

01 Oct, 2025 5 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
- Quick Fix: fix Qwen3-VL launch failure caused by MRotaryEmbedding arg (#10985) · 61305291
  yhyang201 authored Oct 01, 2025
  
  61305291
- [Auto Sync] Update base_grammar_backend.py, xgrammar_backen... (20250930) (#11115) · 5dddb331
  Lianmin Zheng authored Sep 30, 2025
```
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Sehoon Kim <sehoon@x.ai>
```
  5dddb331
- Organize spec-related data structures (#10735) · 73d4a5f8
  Liangsheng Yin authored Oct 01, 2025
  
  73d4a5f8
- fix: KimiK2Detector Improve tool call ID parsing with regex (#10972) · 1193f131
  Xinyuan Tong authored Sep 30, 2025
  
  1193f131
30 Sep, 2025 15 commits
- Feature/make PEFT adapter module format compatibile (#11080) · 84a9f5d6
  Chenxi Li authored Sep 30, 2025
  
  84a9f5d6
- [router][bugfix] Fix input_logprobs handling with None value and `logprob_start_len = -1` (#11113) · 8ce830a8
  Chang Su authored Sep 30, 2025
  
  8ce830a8
- Support Dots.ocr model (#11071) · fb367acf
  qrskannbara authored Oct 01, 2025
  
  fb367acf
- Fix DSR1 accuracy for flashinfer_trtllm MoE with FP8 quantization (#11081) · a6cc86df
  Trevor Morris authored Sep 30, 2025
  
  a6cc86df
- [CPU] Adding Memory Capacity Acquisition Functionality (#11102) · 229d2b95
  Zaili Wang authored Sep 30, 2025
  
  229d2b95
- [Eval] Add `--repeat` in `run_eval` (#11101) · 9710f718
  Liangsheng Yin authored Sep 30, 2025
  
  9710f718
- Fix eagle radix cache (#10846) · 91847e38
  Ke Bao authored Sep 30, 2025
  
  91847e38
- [router][grpc-server] Fix gRPC server shutdown (#11094) · 5a290a56
  Simo Lin authored Sep 30, 2025
  
  5a290a56
- [Auto Sync] Update fused_moe_triton_config.py (20250930) (#11099) · 1237aa19
  Lianmin Zheng authored Sep 30, 2025
```
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Cheng Wan <54331508+ch-wan@users.noreply.github.com>
```
  1237aa19
- [Fix] Resolve performance drop in speculative decoding aiter backend (#11087) · 59911195
  jacky.cheng authored Sep 30, 2025
  
  59911195
- Fix spec filter batch when target extend (#10991) · 424591d5
  Ke Bao authored Sep 30, 2025
  
  424591d5
- Enable optional FP32 compute for LM Head (#10729) · d17986f8
  narutolhy authored Sep 29, 2025
```
Thanks to MiniMax Team and Chenyang Zhao's support.
```
  d17986f8
- [model] added support for w8a8int8 used by neuralmagic/Qwen2-0.5B-Ins… (#9642) · 8831c55c
  DevashishLal-CB authored Sep 29, 2025
  
  8831c55c
- 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
- [Profile] dump memory trace when cuda graph profile is enabled (#11083) · 6535fda1
  Cheng Wan authored Sep 29, 2025
  
  6535fda1
29 Sep, 2025 12 commits
- feat(reasoning): improve enable thinking from request (#10875) · 3713eb61
  Jimmy authored Sep 30, 2025
  
  3713eb61
- [router][grpc] Add logprobs support to router (#11082) · 5937a56d
  Chang Su authored Sep 29, 2025
  
  5937a56d
- fix: fp8 mllama4 without vision modules being quantized (#10611) · 9de1320b
  Mick authored Sep 30, 2025
  
  9de1320b
- 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
- chore: upgrade sgl-kernel 0.3.13 (#11056) · 3a641d90
  Yineng Zhang authored Sep 29, 2025
  
  3a641d90
- chore: upgrade cutedsl 4.2.1 (#11054) · 5942fdb4
  Yineng Zhang authored Sep 29, 2025
  
  5942fdb4
- Fix sgl-kernel benchmark dead code (#11022) · 11965b0d
  Xiaoyu Zhang authored Sep 29, 2025
  
  11965b0d
- Fix gemma 3 launch with `transformers:` the error: `AttributeError:... · 71959545
  Vincent Zhong authored Sep 29, 2025
```
Fix gemma 3 launch with `transformers:` the error: `AttributeError: 'TransformersForCausalLM' object has no attribute 'tp_size'` (#9614)
```
  71959545
- [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
- fix: check if weights are already local before downloading (#11015) · 43fa9f22
  Mick authored Sep 29, 2025
  
  43fa9f22
28 Sep, 2025 8 commits
- fix: show failed models in nightly ci (#10986) · 2e763398
  Mick authored Sep 29, 2025
  
  2e763398
- Update GLM-4.5 Model Doc (#11017) · abb67815
  Yuxuan Zhang authored Sep 29, 2025
  
  abb67815
- Fix FusedSetKVBufferArg in RotaryEmbedding (#11003) · 07440f5f
  Lianmin Zheng authored Sep 28, 2025
  
  07440f5f
- [HiCache] bug: fix mooncake store batch set v1 (#11013) · 9816989b
  Teng Ma authored Sep 28, 2025
  
  9816989b
- prepare for sglang+verl (#10555) · 2a9d995c
  lbk-sys authored Sep 28, 2025
```
Signed-off-by: lbk-sys <hello_lbk@163.com>
Signed-off-by: libaokui <libaokui@hauwei.com>
Co-authored-by: libaokui <libaokui@hauwei.com>
Co-authored-by: Even Zhou <even.y.zhou@outlook.com>
```
  2a9d995c
- [bugfix]Add empty_context import to two_batch_overlap.py (#10964) · a9050b5c
  wejoncy authored Sep 28, 2025
  
  a9050b5c
- Fix CI failure of TypeError: RotaryEmbedding.forward_cpu() got an unexpected... · 66face35
  YanbingJiang authored Sep 28, 2025
```
Fix CI failure of TypeError: RotaryEmbedding.forward_cpu() got an unexpected keyword argument 'fused_set_kv_buffer_arg' (#11009)
```
  66face35
- [router] fix chat template loading and tokenizer path (#10999) · 5519766a
  Simo Lin authored Sep 27, 2025
  
  5519766a