Commits · fa3592cfeb9d7d008e31719fdf68f3dc7d5ad9dd · change / sglang

08 Jun, 2025 2 commits
- rebase h20 fused_moe config (#6966) · fa3592cf
  Xiaoyu Zhang authored Jun 08, 2025
  
  fa3592cf
- Revert "Fuse routed scaling factor in topk_reduce kernel (#6220)" (#6968) · 1fb76ebb
  Yineng Zhang authored Jun 07, 2025
  
  1fb76ebb
07 Jun, 2025 1 commit
- Fuse routed scaling factor in topk_reduce kernel (#6220) · 515ef4fa
  Xiaoyu Zhang authored Jun 08, 2025
  
  515ef4fa
04 Jun, 2025 2 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
- [Refactor] Rename `n_share_experts_fusion` as `num_fused_shared_experts` (#6735) · 8a548052
  Cheng Wan authored Jun 03, 2025
  
  8a548052
28 May, 2025 2 commits
- fix log_info_on_rank0 error when run benchmark (#6260) · 07610353
  Xiaoyu Zhang authored May 28, 2025
  
  07610353
- Refine pre_reorder_triton_kernel slightly to improve performance (#6627) · c087ddd6
  Yuan Luo authored May 28, 2025
```
Co-authored-by: luoyuan.luo <luoyuan.luo@antgroup.com>
```
  c087ddd6
12 May, 2025 3 commits
- Support tuning moe for llama 4 model (#6042) · ef8ec07b
  fzyzcjy authored May 13, 2025
  
  ef8ec07b
- Revert "fix some typos" (#6244) · e8e18dcd
  Lianmin Zheng authored May 12, 2025
  
  e8e18dcd
- fix some typos (#6209) · d738ab52
  applesaucethebun authored May 12, 2025
```
Co-authored-by: Brayden Zhong <b8zhong@uwaterloo.ca>
```
  d738ab52
11 May, 2025 1 commit
- Replace time.time() to time.perf_counter() for benchmarking. (#6178) · 6e2da515
  Lifu Huang authored May 11, 2025
```
Signed-off-by: Lifu Huang <lifu.hlf@gmail.com>
```
  6e2da515
29 Apr, 2025 1 commit
- simplify fused_moe config logging (#5801) · 1cc32603
  Xiaoyu Zhang authored Apr 29, 2025
  
  1cc32603
28 Apr, 2025 3 commits
- add fused moe config for qwen3moe fp8/bf16 (#5849) · a0251a3f
  Yi Zhang authored Apr 29, 2025
  
  a0251a3f
- fused moe triton tuning script support qwen3 (#5842) · e132cba2
  Xiaoyu Zhang authored Apr 29, 2025
  
  e132cba2
- feat: Add fused moe triton config for qwen3 moe on h100 (#5833) · 0045f4b2
  XinyuanTong authored Apr 28, 2025
  
  0045f4b2
20 Apr, 2025 1 commit
- Minor update for ROCm variable style (#5562) · c555d794
  Zhaoyi Li authored Apr 20, 2025
  
  c555d794
15 Apr, 2025 1 commit
- Add A800 shared experts fused MoE kernel tuning configs for DeepSeek V3/R1 (#5368) · 61e7c4dd
  lambert0312 authored Apr 15, 2025
  
  61e7c4dd
12 Apr, 2025 1 commit
- refine fused_moe tuning docs (#5294) · 3e4794aa
  Xiaoyu Zhang authored Apr 13, 2025
  
  3e4794aa
04 Apr, 2025 1 commit
- Add DeepSeek V3/R1 shared experts fusion (#4918) · 924ca7c9
  Xiaoyu Zhang authored Apr 04, 2025
  
  924ca7c9
22 Mar, 2025 1 commit
- Correcting default configuration when benchmarking fused_moe (#4665) · 38f25e87
  penguin_wwy authored Mar 22, 2025
  
  38f25e87
12 Mar, 2025 1 commit
- refine sgl_moe_align_block_size_benchmark (#4327) · 7130a7ce
  Xiaoyu Zhang authored Mar 12, 2025
  
  7130a7ce
11 Mar, 2025 1 commit
- Add A100 tuning configs for DeepSeek R1/V3 channel-wise INT8 (#4287) · 6a02b32d
  yych0745 authored Mar 11, 2025
```
Co-authored-by: HandH1998 <1335248067@qq.com>
```
  6a02b32d
03 Mar, 2025 2 commits

Improve code styles (#4021) · 66301e12
Lianmin Zheng authored Mar 03, 2025

66301e12

Support penalty in overlap mode; return logprob with chunked prefill; improve... · ac238727

Lianmin Zheng authored Mar 03, 2025


Support penalty in overlap mode; return logprob with chunked prefill; improve benchmark scripts (#3988)
Co-authored-by: SangBin Cho <rkooo567@gmail.com>
Co-authored-by: dhou-xai <dhou@x.ai>
Co-authored-by: Hanming Lu <hanming_lu@berkeley.edu>

ac238727

01 Mar, 2025 1 commit
- Revert "[MOE] enable efficient moe_alignment multi-blocks execution (3x~6x)" (#3982) · 18bb216c
  Chayenne authored Feb 28, 2025
  
  18bb216c
28 Feb, 2025 1 commit
- [MOE] enable efficient moe_alignment multi-blocks execution (3x~6x) (#3613) · 1c96fa86
  yiakwy-xpu-ml-framework-team authored Feb 28, 2025
  
  1c96fa86
27 Feb, 2025 1 commit
- Tuning Script for Feature DeepSeek V3/R1 INT8 Quantization (block-wise) (#3922) · b0df5d24
  laixin authored Feb 27, 2025
```
Co-authored-by: sleepcoo <sleepcoo@gmail.com>
```
  b0df5d24
18 Feb, 2025 1 commit
- [ROCm] Optimal MOE Tuning for AMD Radeon Graphics (#3567) · ddf39d3f
  yigex authored Feb 18, 2025
  
  ddf39d3f
10 Feb, 2025 1 commit
- refine some typo (#3473) · 2f47d710
  Xiaoyu Zhang authored Feb 10, 2025
  
  2f47d710
08 Feb, 2025 3 commits
- fix EAGLE 2 non greedy case (#3407) · fad315cb
  Yineng Zhang authored Feb 09, 2025
```
Co-authored-by: Ying Sheng <sqy1415@gmail.com>
```
  fad315cb
- Update fused_moe's benchmark (#3346) · 849f58d6
  GaoYuYang authored Feb 08, 2025
  
  849f58d6
- [BUG] fix moe benchmark when bs*seq is small (#3382) · 64480df4
  yiakwy-xpu-ml-framework-team authored Feb 08, 2025
  
  64480df4
06 Feb, 2025 2 commits
- optimize moe_align_kernel cuda (#3347) · cdae77b0
  Xiaoyu Zhang authored Feb 07, 2025
  
  cdae77b0
- clean moe align block kernel code and add acc test (#3332) · ad349985
  Xiaoyu Zhang authored Feb 06, 2025
  
  ad349985
27 Jan, 2025 1 commit
- add dsv3 mi300 triton config for block scale (#3146) · 351a72d4
  yigex authored Jan 27, 2025
  
  351a72d4
26 Jan, 2025 1 commit
- Use torch.compile for scaling penalty (#3133) · 27acf63b
  Lianmin Zheng authored Jan 25, 2025
  
  27acf63b
20 Jan, 2025 1 commit
- fix moe align blocks benchmark (#3003) · 10bfce71
  yiakwy-xpu-ml-framework-team authored Jan 20, 2025
  
  10bfce71
13 Jan, 2025 1 commit
- Sampling penalties memory interface (#2870) · d08c77c4
  Xiaoyu Zhang authored Jan 13, 2025
  
  d08c77c4
07 Jan, 2025 2 commits
- Misc fix for min_p_sampling, --cuda-graph-bs (#2761) · bdc1acf6
  Lianmin Zheng authored Jan 07, 2025
  
  bdc1acf6
- add benchmark_moe_align_blocks (#2767) · 380930a9
  Xiaoyu Zhang authored Jan 07, 2025
  
  380930a9