Xiaoyu Zhang 8b5f83ed3b reduce torch.zeros overhead in moe align block size kernel (#6369)

2025-06-07 02:47:36 -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

Cutlass MLA: Disable split kv due to https://github.com/NVIDIA/cutlass/issues/2274 (#6101)

2025-05-08 18:44:30 -07:00

cpu

Add CPU optimized kernels for topk and rope fusions (#6456)

2025-06-02 17:37:34 -07:00

cutlass_extensions

sgl-kernel use cutlass latest version for fp8 blockwise gemm (#5207)

2025-04-09 11:47:04 -07:00

elementwise

[Feat] Update sgl-kernel flashinfer to latest main version (#5500)

2025-04-17 12:43:23 -07:00

gemm

[1/2] Add Kernel support for Cutlass based Fused FP4 MoE (#6093)

2025-06-02 13:48:03 -07:00

grammar

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

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

moe

reduce torch.zeros overhead in moe align block size kernel (#6369)

2025-06-07 02:47:36 -07:00

speculative

Add typo checker in pre-commit (#6179)

2025-05-11 12:55:00 +08:00

common_extension.cc

[EP] Add cuda kernel for moe_ep_post_reorder (#6837)

2025-06-05 00:33:47 -07:00

flash_extension.cc

[Fix] fix fa3 build at cu118 (#5036)

2025-04-03 11:52:35 -07:00

torch_extension_rocm.cc

update variable naming and comments for rocm (#5299)

2025-04-11 23:15:05 -07:00