Ke Bao a3398d8478 Optimize moe align block size kernel (#7794)

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

..

cutlass_moe/w4a8

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

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

marlin_moe_wna16

[1/n] apply wna16marlin kernel in moe weight only quantization (#7683)

2025-07-01 23:21:25 -07:00

cutlass_moe_helper.cu

…

ep_moe_reorder_kernel.cu

…

ep_moe_silu_and_mul_kernel.cu

…

fp8_blockwise_moe_kernel.cu

Optimize Hopper CUTLASS FP8 Blockwise Grouped GEMM Kernel in Small K Scenario (#7782)

2025-07-04 22:25:49 -07:00

moe_align_kernel.cu

Optimize moe align block size kernel (#7794)

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

moe_fused_gate.cu

…

moe_topk_softmax_kernels.cu

[optimize] fuse renormalize into moe_topk_softmax (#7744)

2025-07-03 12:42:44 -07:00

nvfp4_blockwise_moe.cu

…

prepare_moe_input.cu

fix: fix apply_shuffle_mul_sum (#7444)

2025-07-04 23:23:30 -07:00