root c5600bc8ae Add decode tiles (b16x32, b32x32) to pagedkv_prefill codegen with max_seqlen_q dispatch ...

Made-with: Cursor

2026-04-01 18:30:06 +00:00

..

__init__.py

…

fmha_batch_prefill.py

Add unified attention (42_unified_attention)

2026-04-01 16:39:15 +00:00

fmha_bwd.py

[rocm-libraries] ROCm/rocm-libraries#5174 (commit a358a21)

2026-03-13 06:14:31 +00:00

fmha_fwd_appendkv.py

…

fmha_fwd_splitkv.py

Split-KV codegen: dual-tile dispatch and head-merge for hdim=64

2026-04-01 16:24:25 +00:00

fmha_fwd.py

Split-KV codegen: dual-tile dispatch and head-merge for hdim=64

2026-04-01 16:24:25 +00:00

fmha_pagedkv_prefill.py

Add decode tiles (b16x32, b32x32) to pagedkv_prefill codegen with max_seqlen_q dispatch

2026-04-01 18:30:06 +00:00