Commits · bdd17998e60f6151bf6d1f7a8e1cadfb73a0aaa4 · change / sglang

29 Apr, 2025 1 commit
- cutlass 3.9 supported to improve fp8_blockwise_gemm (#5820) · 5bb0accb
  Xiaoyu Zhang authored Apr 29, 2025
  
  5bb0accb
28 Apr, 2025 1 commit
- [Feat] QWen-1M context support[1/2]: Update block sparse attention backend utils kernel (#5847) · ee71ed8a
  PGFLMG authored Apr 29, 2025
```
Co-authored-by: sighingnow <sighingnow@gmail.com>
```
  ee71ed8a
23 Apr, 2025 3 commits
- fix: only compile ApplyTokenBitmaskInplace cu124+ (#5686) · ce4ecba4
  Yineng Zhang authored Apr 23, 2025
  
  ce4ecba4
- fix sgl-kernel unit tests (#5666) · 15fabcc0
  Yineng Zhang authored Apr 23, 2025
  
  15fabcc0
- [1/2] Add FP8 Blockscale MoE CUTLASS kernel for Blackwell (#5281) · e62c4955
  Elfie Guo authored Apr 22, 2025
  
  e62c4955
18 Apr, 2025 1 commit
- Sgl kernel fused_moe_gate support n_shared_experts (#5440) · 8e09b370
  Xiaoyu Zhang authored Apr 18, 2025
  
  8e09b370
17 Apr, 2025 1 commit
- [Feat] Update sgl-kernel flashinfer to latest main version (#5500) · c08a717c
  PGFLMG authored Apr 18, 2025
```
Co-authored-by: zhyncs <me@zhyncs.com>
```
  c08a717c
15 Apr, 2025 2 commits
- bugfix: fix merge_state_v2 cuda graph (#5419) · 12ef7e3b
  DefTruth authored Apr 16, 2025
  
  12ef7e3b
- kernel: support slightly faster merge_state_v2 cuda kernel (#5381) · 388e15c0
  DefTruth authored Apr 15, 2025
  
  388e15c0
13 Apr, 2025 1 commit
- feat: adapt merge_state (#5337) · b62e7e99
  Yineng Zhang authored Apr 12, 2025
  
  b62e7e99
12 Apr, 2025 4 commits
- fix: solve cu118 issue for cutlass mla (#5331) · 812e82f3
  Yineng Zhang authored Apr 12, 2025
  
  812e82f3
- [Feat] Add sparse attn to sgl-kernel (#5327) · 4879e50c
  PGFLMG authored Apr 13, 2025
  
  4879e50c
- update variable naming and comments for rocm (#5299) · 3c9740d2
  Zhaoyi Li authored Apr 12, 2025
  
  3c9740d2
- Blackwell Cutlass MLA kernel (#5142) · f65b8d5c
  Trevor Morris authored Apr 11, 2025
  
  f65b8d5c
11 Apr, 2025 1 commit
- fix: remove cublas_grouped_gemm (#5307) · 136b8e6a
  Yineng Zhang authored Apr 11, 2025
  
  136b8e6a
10 Apr, 2025 2 commits
- Fix deepseek-v3 with torch.compile in PyTorch 2.6. (#5213) · 76f44c2a
  Richard Zou authored Apr 10, 2025
  
  76f44c2a
- reduce moe_align_block_size_kernel small batch mode overhead (#5086) · f730362e
  Xiaoyu Zhang authored Apr 10, 2025
  
  f730362e
09 Apr, 2025 1 commit
- sgl-kernel use cutlass latest version for fp8 blockwise gemm (#5207) · ebf495f0
  Yi Zhang authored Apr 10, 2025
  
  ebf495f0
08 Apr, 2025 1 commit

Add optimized native kernels in sgl-kernel (#5150) · a73c4df4

Ma Mingfei authored Apr 09, 2025


Co-authored-by: Chunyuan WU <chunyuan.wu@intel.com>
Co-authored-by: YanbingJiang <yanbing.jiang@intel.com>
Co-authored-by: blzheng <beilei.zheng@intel.com>

a73c4df4

05 Apr, 2025 1 commit
- sgl-kernel transfer custom allreduce from trt kernel to vllm kernel (#5079) · bcbbf519
  Yi Zhang authored Apr 06, 2025
  
  bcbbf519
03 Apr, 2025 2 commits
- [Fix] fix fa3 build at cu118 (#5036) · b8b6008f
  yinfan98 authored Apr 04, 2025
  
  b8b6008f
- [sgl-kernel] per token group quant support COLUMN MAJOR (#4817) · 2c8fd993
  Xiaoyu Zhang authored Apr 03, 2025
  
  2c8fd993
31 Mar, 2025 2 commits
- [Build] Fix cuda12.8 build error in nvfp4_scaled_mm_kernels.cu (#4953) · ee47a6c1
  Yuhong Guo authored Apr 01, 2025
  
  ee47a6c1
- [Fix] revert clean m.def for cudagraph (#4944) · c7457191
  yinfan98 authored Mar 31, 2025
  
  c7457191
30 Mar, 2025 3 commits
- [feat] add fa3 in sgl-kernel (#4902) · 37c66ec8
  yinfan98 authored Mar 31, 2025
```
Co-authored-by: Sleepcoo <Sleepcoo@gmail.com>
```
  37c66ec8
- fix bmm fp8 (#4926) · 195a09f5
  Yineng Zhang authored Mar 30, 2025
  
  195a09f5
- [Misc] Clean m.def and add Development Tips (#4890) · 0d7fe866
  yinfan98 authored Mar 30, 2025
  
  0d7fe866
29 Mar, 2025 2 commits
- Add deepseek style fused moe group gate selection kernel (#4530) · 45dcfc2e
  Qingquan Song authored Mar 29, 2025
  
  45dcfc2e
- fix sgl-kernel cu118 build (#4872) · ec3ee028
  Yineng Zhang authored Mar 28, 2025
  
  ec3ee028
27 Mar, 2025 1 commit

support cmake for sgl-kernel (#4706) · 8bf6d7f4

Yineng Zhang authored Mar 27, 2025


Co-authored-by: hebiao064 <hebiaobuaa@gmail.com>
Co-authored-by: yinfan98 <1106310035@qq.com>

8bf6d7f4

26 Mar, 2025 1 commit
- Fix shared memory OOM on sm86 GPUs. (#4797) · 45fdf1f7
  Yi Pan authored Mar 27, 2025
  
  45fdf1f7
25 Mar, 2025 1 commit
- Support FP4 gemm (1/2) (#3899) · e9f8e423
  Trevor Morris authored Mar 24, 2025
  
  e9f8e423
24 Mar, 2025 2 commits
- [Quant Kernel] refactored per token group quant fp8 to support int8 up-to 2x faster (#4396) · 65c24c28
  Chunan Zeng authored Mar 23, 2025
  
  65c24c28
- [ROCm] Enable MTP (NextN) on AMD GPU (#4631) · af6535e7
  Alex Sun authored Mar 24, 2025
  
  af6535e7
22 Mar, 2025 2 commits
- [1/3] fix dsv3 awq issue (#4556) · 321ab756
  AniZpZ authored Mar 22, 2025
```
Co-authored-by: leoneo <1320612015@qq.com>
```
  321ab756
- Speed up per token and per tensor quant by 15% (#4639) · 6a384d5c
  Chunan Zeng authored Mar 22, 2025
  
  6a384d5c
20 Mar, 2025 1 commit
- Support fp8 gemm for blackwell (#4558) · ad4e58bf
  Shu Wang authored Mar 20, 2025
  
  ad4e58bf
17 Mar, 2025 2 commits
- Support serving DeepSeek-R1-Channel-INT8 with 32 L40S. (#4418) · 75b65648
  Wenbo Yang authored Mar 17, 2025
  
  75b65648
- [ROCm] enable moe topk softmax in amd (#4448) · 9b8333d9
  yiakwy-xpu-ml-framework-team authored Mar 17, 2025
  
  9b8333d9
16 Mar, 2025 1 commit
- fix custom allreduce performance/accuracy problem (#4477) · 25e1816e
  Yi Zhang authored Mar 17, 2025
  
  25e1816e