Commits · 8b5f83ed3b7d2a49ad5c5cd5aa61c5d502f47dbc · zhaoyu6 / sglang

"lm_eval/tasks/squadv2/task.py" did not exist on "fe4a1efdee679e8ec72c1dcc30a90caf8ed4b2ed"

07 Jun, 2025 13 commits
- reduce torch.zeros overhead in moe align block size kernel (#6369) · 8b5f83ed
  Xiaoyu Zhang authored Jun 07, 2025
  
  8b5f83ed
- Add triton version as a fused_moe_triton config search key to avoid performace... · 2a413829
  Xiaoyu Zhang authored Jun 07, 2025
```
Add triton version as a fused_moe_triton config search key to avoid performace decrease in different Triton version (#5955)
```
  2a413829
- Tiny re-introduce profile id logging (#6912) · d5c097a2
  fzyzcjy authored Jun 07, 2025
  
  d5c097a2
- [Bugfix] pipeline parallelism and Eagle Qwen2 (#6910) · 9736cd3b
  Swipe4057 authored Jun 07, 2025
  
  9736cd3b
- Minor compile fused topk (#6944) · 2f715f51
  fzyzcjy authored Jun 07, 2025
  
  2f715f51
- chore: bump sgl-kernel v0.1.6 (#6943) · d664ca18
  Yineng Zhang authored Jun 07, 2025
  
  d664ca18
- [sgl-kernel] update deepgemm (#6942) · 22fe7878
  JieXin Liang authored Jun 07, 2025
  
  22fe7878
- Add triton fused moe kernel config for E=257 on B200 (#6939) · c4ffbeca
  Baizhou Zhang authored Jun 06, 2025
  
  c4ffbeca
- [fix] logical_to_all_physical_map index 256 is out of bounds in EP parallel. (#6767) · f8eaaab8
  miter authored Jun 07, 2025
```
Signed-off-by: miter <miterv@outlook.com>
```
  f8eaaab8
- [Refactor] image data process in bench_serving (#6879) · 697b0f71
  Xinyuan Tong authored Jun 06, 2025
```
Signed-off-by: Xinyuan Tong <justinning0323@outlook.com>
```
  697b0f71
- [PD] Optimize transfer queue forward logic for dummy rank (#6922) · 132dad87
  shangmingc authored Jun 07, 2025
```
Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com>
```
  132dad87
- Sync the changes on cuda graph runners (#6932) · 60fdad7c
  Lianmin Zheng authored Jun 06, 2025
  
  60fdad7c
- Tiny support customize DeepEP max dispatch tokens per rank (#6934) · 61ce91ed
  fzyzcjy authored Jun 07, 2025
  
  61ce91ed
06 Jun, 2025 3 commits
- Fix a bug in abort & Improve docstrings for abort (#6931) · e6b7053b
  Lianmin Zheng authored Jun 06, 2025
  
  e6b7053b
- [Feature] Support Flashinfer fmha on Blackwell (#6930) · 5f91c825
  Jianan Ji authored Jun 06, 2025
  
  5f91c825
- AITER backend extension and workload optimizations (#6838) · b819381f
  HAI authored Jun 05, 2025
```
Co-authored-by: wunhuang <wunhuang@amd.com>
Co-authored-by: Hubert Lu <Hubert.Lu@amd.com>
```
  b819381f
05 Jun, 2025 15 commits
- [CPU] enable CI for PRs, add Dockerfile and auto build task (#6458) · 562f279a
  Zaili Wang authored Jun 06, 2025
```
Co-authored-by: diwei sun <diwei.sun@intel.com>
Co-authored-by: Yineng Zhang <me@zhyncs.com>
```
  562f279a
- bugfix(OAI): Fix image_data processing for jinja chat templates (#6877) · 8b247489
  Chang Su authored Jun 05, 2025
  
  8b247489
- [CUTLASS-FP4-MOE] Introduce CutlassMoEParams class for easy initialization of... · 0df6765c
  Pavani Majety authored Jun 05, 2025
```
[CUTLASS-FP4-MOE]  Introduce CutlassMoEParams class for easy initialization of Cutlass Grouped Gems Metadata (#6887)
Signed-off-by: Pavani Majety <pmajety@nvidia.com>
```
  0df6765c
- Use deepgemm instead of triton for fused_qkv_a_proj_with_mqa (#6890) · 35b65cf0
  fzyzcjy authored Jun 06, 2025
  
  35b65cf0
- [PD] Fix potential perf spike caused by tracker gc and optimize doc (#6764) · dd1012fc
  shangmingc authored Jun 06, 2025
```
Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com>
```
  dd1012fc
- oai: fix openAI client error with single request via batch api (#6170) · 44aab7f9
  Ravi Theja authored Jun 05, 2025
```
Co-authored-by: Ravi Theja Desetty <ravitheja@Ravis-MacBook-Pro.local>
```
  44aab7f9
- [EP] Add cuda kernel for moe_ep_post_reorder (#6837) · 43baba64
  Yuan Luo authored Jun 05, 2025
```
Co-authored-by: luoyuan.luo <luoyuan.luo@antgroup.com>
```
  43baba64
- Support Blackwell DeepEP docker images (#6868) · 0166403c
  fzyzcjy authored Jun 05, 2025
  
  0166403c
- Tiny allow profiler API to auto create directory (#6865) · bcf66ef3
  fzyzcjy authored Jun 05, 2025
  
  bcf66ef3
- Support layerwise rebalancing experts (#6851) · 0de5e7d4
  fzyzcjy authored Jun 05, 2025
  
  0de5e7d4
- Tiny update error hints (#6846) · 72a110f6
  fzyzcjy authored Jun 05, 2025
  
  72a110f6
- Fix Qwen3MoE missing token padding optimization (#6820) · 5aff1e93
  fzyzcjy authored Jun 05, 2025
  
  5aff1e93
- support 1 shot allreduce in 1-node and 2-node using mscclpp (#6277) · 8e3797be
  zyksir authored Jun 05, 2025
  
  8e3797be
- Support LoRA in TestOpenAIVisionServer and fix fused kv_proj loading bug. (#6861) · 4474eaf5
  Lifu Huang authored Jun 04, 2025
  
  4474eaf5
- Fix one missing arg in DeepEP (#6878) · 499f5e62
  Cheng Wan authored Jun 04, 2025
  
  499f5e62
04 Jun, 2025 8 commits
- Set `num_fused_shared_experts` as `num_shared_experts` when shared_experts... · 81964328
  Cheng Wan authored Jun 04, 2025
```
Set `num_fused_shared_experts` as `num_shared_experts` when shared_experts fusion is not disabled (#6736)
```
  81964328
- feat: add dp-rank to KV events (#6852) · f0f84975
  ishandhanani authored Jun 04, 2025
  
  f0f84975
- Decoder-only Scoring API (#6460) · 3f1e4339
  Chanh Nguyen authored Jun 04, 2025
```
Co-authored-by: Chanh Nguyen <cnguyen@linkedin.com>
```
  3f1e4339
- [Refactor] Multimodal data processing for VLM (#6659) · cf9815ba
  Xinyuan Tong authored Jun 04, 2025
```
Signed-off-by: Xinyuan Tong <justinning0323@outlook.com>
```
  cf9815ba
- fix ep_moe_reorder kernel bugs (#6858) · bd75690f
  Xiaoyu Zhang authored Jun 04, 2025
```
Co-authored-by: JieXin Liang <Alcanderian@users.noreply.github.com>
```
  bd75690f
- [fix] recover auto-dispatch for rmsnorm and rope (#6745) · 180ff5ee
  JieXin Liang authored Jun 04, 2025
  
  180ff5ee
- [FEAT] Add transformers backend support (#5929) · 37f15475
  Marc Sun authored Jun 04, 2025
  
  37f15475
- [Refactor] Rename `n_share_experts_fusion` as `num_fused_shared_experts` (#6735) · 8a548052
  Cheng Wan authored Jun 03, 2025
  
  8a548052
03 Jun, 2025 1 commit
- Minor add metrics to expert location updater (#6816) · b6d0ce9f
  fzyzcjy authored Jun 03, 2025
  
  b6d0ce9f