Zzz9990 ece38b9d7a [VLLM V1] Add chunked prefill for FA to pass seq with small seqlen_q (#2221) ...

* fix splitkv compiler issue since lse is used to select kernel instances

* bypass seqlen == 1

* add chunked prefill into mha varlen

This reverts commit aa9847e42d.

* skip compile when receipt 2-4 and add comments

* fix

---------

Co-authored-by: fsx950223 <fsx950223@outlook.com>

2025-05-26 19:17:18 +08:00

..

fmha_batch_prefill_kernel.hpp

[CK_TILE] fMHA batch_prefill block index & logits soft-capping optimizations (#2198)

2025-05-16 15:14:46 +08:00

fmha_bwd_kernel.hpp

Sync the kname with instance name (#1989)

2025-03-20 00:06:45 +08:00

fmha_fwd_appendkv_kernel.hpp

Remove using partitioner for all fmha kernels (#1778)

2024-12-29 14:29:56 +08:00

fmha_fwd_appendkv_tile_partitioner.hpp

[CK_TILE] Add PagedAttention kernels (#1387)

2024-08-28 20:50:43 +08:00

fmha_fwd_kernel.hpp

[VLLM V1] Add chunked prefill for FA to pass seq with small seqlen_q (#2221)

2025-05-26 19:17:18 +08:00

fmha_fwd_splitkv_combine_kernel.hpp

Sync the kname with instance name (#1989)

2025-03-20 00:06:45 +08:00

fmha_fwd_splitkv_kernel.hpp

[CK_TILE] Add logits soft-capping & customization support to the FMHA forward kernel/pipelines (#2163)

2025-05-13 12:19:25 +08:00