| .. |
|
fmha
|
chore: Improve documentation of Kv_block_array (#5765)
|
2025-07-05 22:25:27 +02:00 |
|
convert.cu
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_attention_demo_bert_params.h
|
use cu for fmha_v2 (#4694)
|
2025-06-15 18:40:44 +08:00 |
|
fused_multihead_attention_kernel_1xN_multi_cta.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_attention_kernel_1xN_noloop.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_attention_kernel_1xN.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_attention_kernel_2x2.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_attention_kernel_4x1_hopper_noloop.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_attention_kernel_4x1_hopper.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_attention_kernel_4xN_hopper_noloop.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_attention_kernel_4xN_hopper.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_attention_kernel.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_attention_utils.h
|
[Feat] add chunked-attention kernels on Hopper (for llama4) (#4291)
|
2025-05-19 09:57:10 -07:00 |
|
fused_multihead_attention.cpp
|
fix: fmha_v2 compilation (#4659)
|
2025-05-27 17:39:39 +08:00 |
|
fused_multihead_attention.h
|
use cu for fmha_v2 (#4694)
|
2025-06-15 18:40:44 +08:00 |
|
fused_multihead_cross_attention_kernel_1xN_noloop.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_cross_attention_kernel_1xN.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_cross_attention.cpp
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_cross_attention.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
fused_multihead_flash_attention_kernel_noloop_tiled.h
|
[Feat] add chunked-attention kernels on Hopper (for llama4) (#4291)
|
2025-05-19 09:57:10 -07:00 |
|
fused_multihead_flash_attention_kernel_noloop.h
|
[Feat] add chunked-attention kernels on Hopper (for llama4) (#4291)
|
2025-05-19 09:57:10 -07:00 |
|
fused_multihead_flash_attention_kernel.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
softmax_bf16.cu
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
softmax_fp8.cu
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
softmax_fp16.cu
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
softmax_fp32.cu
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
softmax_impl.h
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
|
softmax_int8.cu
|
infra: open source fmha v2 kernels (#4185)
|
2025-05-15 10:56:34 +08:00 |
