Yi DING 8e1eb0c1ee [CK_TILE] FMHA BWD Decode Pipeline (#2643) ...

* Fix distr

* Duplicate block_fmha_bwd_dq_dk_dv_pipeline_trload_kr_ktr_vr

* decode 16x16 o2

2025-08-12 17:02:52 +08:00

..

fmha_batch_prefill_kernel.hpp

[CK_TILE] Add missing parameter 'min_seqlen_q' to the FMHA fwd kernel MakeKargs() interface (#2403)

2025-06-25 15:19:21 +08:00

fmha_bwd_kernel.hpp

[CK_TILE] FMHA BWD Decode Pipeline (#2643)

2025-08-12 17:02:52 +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

reorder grid dim schedule (#2533)

2025-07-26 02:46:55 +08:00

fmha_fwd_pagedkv_kernel.hpp

Add separate mask checking for scope [aligned_physical_seqlen_k_start, physical_seqlen_k_end) (#2487)

2025-07-11 18:14:47 +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] For FMHA forward kernels, assign block indices reversely if using mask (#2209)

2025-05-27 10:58:58 +08:00