Commits · a2cb5913a00965a8bb33f3986710adaaf1828a8e · zhaoyu6 / sglang

"lightx2v/models/schedulers/vscode:/vscode.git/clone" did not exist on "b50498fab8f855c1504f13baee4ba8877f3d4aa9"

02 Jun, 2025 1 commit
- Add draft extend CUDA graph for flashinfer backend (#6805) · a2cb5913
  Ke Bao authored Jun 02, 2025
  
  a2cb5913
01 Jun, 2025 1 commit
- Update CI tests for Llama4 models (#6421) · c6a0cacc
  Ravi Theja authored Jun 01, 2025
  
  c6a0cacc
31 May, 2025 5 commits
- Improve profiler and integrate profiler in bench_one_batch_server (#6787) · 2d72fc47
  Lianmin Zheng authored May 31, 2025
  
  2d72fc47
- ci: relax test_function_call_required (#6786) · e39bca07
  Chang Su authored May 30, 2025
  
  e39bca07
- Temporarily lower mmlu threshold for triton sliding window backend (#6785) · a2bb8565
  Jianan Ji authored May 30, 2025
  
  a2bb8565
- feat(tool call): Enhance Llama32Detector for improved JSON parsing in non-stream (#6784) · f18b068f
  Chang Su authored May 30, 2025
  
  f18b068f
- update llama4 chat template and pythonic parser (#6679) · 4fac524b
  Chao Yang authored May 30, 2025
```
Co-authored-by: Chang Su <chang.s.su@oracle.com>
```
  4fac524b
30 May, 2025 1 commit
- Support sliding window in triton backend (#6509) · 22630ca2
  Jianan Ji authored May 30, 2025
  
  22630ca2
29 May, 2025 2 commits
- refactor(tool call): Fix BaseFormatDetector tool_index issue and refactor... · c673727e
  Chang Su authored May 29, 2025
```
refactor(tool call): Fix BaseFormatDetector tool_index issue and refactor `parse_streaming_increment` (#6715)
```
  c673727e
- Fix PP for Qwen3 MoE (#6709) · e06b0761
  iLeGend authored May 29, 2025
  
  e06b0761
28 May, 2025 3 commits
- [CI] Fix setup of disaggregation with different tp (#6706) · d63e76f7
  shangmingc authored May 29, 2025
```
Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com>
```
  d63e76f7
- [CI] Fix flaky pp single node test (#6689) · c25231c6
  shangmingc authored May 28, 2025
```
Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com>
```
  c25231c6
- Update amd docker and nightly models. (#6687) · f4a8987f
  Sai Enduri authored May 28, 2025
  
  f4a8987f
27 May, 2025 4 commits
- feat: Add warnings for invalid tool_choice and UTs (#6582) · 41ba767f
  Chang Su authored May 27, 2025
  
  41ba767f
- Add batch test for draft extend (#6672) · f127355a
  Ke Bao authored May 28, 2025
  
  f127355a
- Support gathering expert distribution details (#6665) · 87068b5c
  fzyzcjy authored May 28, 2025
  
  87068b5c
- [CI] update verlengine ci to 4-gpu test (#6007) · 2103b806
  Junrong Lin authored May 28, 2025
  
  2103b806
26 May, 2025 7 commits
- Update nightly thresholds and dependencies. (#6635) · eb8f02dd
  Sai Enduri authored May 26, 2025
  
  eb8f02dd
- Supported precomputed feature for Kimi VL (#6599) · 0d503090
  Lifu Huang authored May 26, 2025
  
  0d503090
- qwen3moe support two batch overlap (#6598) · f9bab3d5
  Yi Zhang authored May 26, 2025
  
  f9bab3d5
- feat: Improve Mistral and Qwen25 function call parsing (#6597) · 16f69b1f
  Chang Su authored May 25, 2025
  
  16f69b1f
- refactor qwen moe code, use communicator to support tp+dp (#6581) · 65f09131
  Yi Zhang authored May 26, 2025
  
  65f09131
- chore: upgrade transformers 4.52.3 (#6575) · 7eb9d8e5
  Yineng Zhang authored May 25, 2025
```
Co-authored-by: Mick <mickjagger19@icloud.com>
```
  7eb9d8e5
- Improve performance of two batch overlap in some imbalanced cases (#6593) · a191a0e4
  fzyzcjy authored May 26, 2025
  
  a191a0e4
25 May, 2025 4 commits
- added support for tied weights in qwen pipeline parallelism (#6546) · 3f23d8cd
  Shenggui Li authored May 25, 2025
  
  3f23d8cd
- Support Phi-4 Multi-Modal (text + vision only) (#6494) · 022012aa
  Lifu Huang authored May 24, 2025
  
  022012aa
- Refactor vlm embedding routine to use precomputed feature (#6543) · 681fdc26
  Xinyuan Tong authored May 24, 2025
```
Signed-off-by: Xinyuan Tong <justinning0323@outlook.com>
```
  681fdc26
- Support overlapping two batches (#4068) · 0d477880
  fzyzcjy authored May 25, 2025
  
  0d477880
24 May, 2025 3 commits
- Fix GPU OOM (#6564) · 7a5e6ce1
  kk authored May 25, 2025
```
Co-authored-by: michael <michael.zhang@amd.com>
```
  7a5e6ce1
- [PD] Support structured output (#6560) · 2d831c6e
  Byron Hsu authored May 23, 2025
  
  2d831c6e
- feat(Tool Calling): Support `required` and specific function mode (#6550) · ed0c3035
  Chang Su authored May 23, 2025
  
  ed0c3035
23 May, 2025 7 commits
- fix: remove content=none test when tool called (#6347) · 9c574585
  Shi Shuai authored May 24, 2025
  
  9c574585
- [PD] Support logprob & Add failure test (#6558) · 8233cc10
  Byron Hsu authored May 23, 2025
  
  8233cc10
- [PD] support spec decode (#6507) · d2e0881a
  Byron Hsu authored May 23, 2025
```
Co-authored-by: SangBin Cho <rkooo567@gmail.com>
```
  d2e0881a
- Update sgl-kernel UTs for activation/topk/norm/rope kernels (#6452) · d8189660
  YanbingJiang authored May 23, 2025
  
  d8189660
- Add fp8 fused_experts kernel for CPU in sgl-kernel and add UT (#6404) · 3ded6235
  Chunyuan WU authored May 23, 2025
  
  3ded6235
- Add fp8 qkv_proj_with_rope kernel for CPU in sgl-kernel and add UT (#6493) · 4ba1eea8
  blzheng authored May 23, 2025
  
  4ba1eea8
- [VLM] Support chunk prefill for VLM (#6355) · 4685fbb8
  Chang Su authored May 22, 2025
```
Co-authored-by: yizhang2077 <1109276519@qq.com>
```
  4685fbb8
22 May, 2025 2 commits
- Support XiaomiMiMo inference with mtp (#6059) · a6ae3af1
  ryang authored May 23, 2025
  
  a6ae3af1
- chore: upgrade sgl-kernel v0.1.4 (#6532) · 0b07c4a9
  Yineng Zhang authored May 22, 2025
  
  0b07c4a9