likesen-alibaba 4a0d19198b Fix bug of deepseek-v3 under DP+EP mode with large batchsize/seqlen (#6449)

2025-07-10 01:19:56 -07:00

..

allreduce

support 1 shot allreduce in 1-node and 2-node using mscclpp (#6277)

2025-06-04 22:11:24 -07:00

attention

[fix] fix cutlass_mla_backend with cuda_graph and add sm_scale for sgl-kernel cutlass_mla (#7184)

2025-06-14 12:45:41 -07:00

cpu

[CPU] [FP8] set SGLANG_CPU_FP8_CVT_FTZ in CMakeLists.txt (#7885)

2025-07-09 01:53:53 -07:00

cutlass_extensions

[1/n]: add cutlass W4A8 moe kernel for hopper architecture (#7772)

2025-07-04 20:50:12 -07:00

elementwise

…

gemm

Fix bug of deepseek-v3 under DP+EP mode with large batchsize/seqlen (#6449)

2025-07-10 01:19:56 -07:00

grammar

[sgl-kernel] fix: fix cu118 compile error (#6123)

2025-05-08 14:26:51 -07:00

kvcacheio

kvcache io kernels and test case (#7382)

2025-06-23 11:58:59 -07:00

moe

Optimize moe align block size kernel (#7794)

2025-07-07 09:20:30 +08:00

speculative

Add treemask mode to build_eagle_tree & release sgl-kernel 0.2.3 (#7756)

2025-07-05 12:17:05 -07:00

common_extension.cc

Add treemask mode to build_eagle_tree & release sgl-kernel 0.2.3 (#7756)

2025-07-05 12:17:05 -07:00

flash_extension.cc

…

torch_extension_rocm.cc

Add treemask mode to build_eagle_tree & release sgl-kernel 0.2.3 (#7756)

2025-07-05 12:17:05 -07:00