ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-14 10:09:41 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
a57f8d8b67cf28547b61dafd2160dffb33b6eaeb
composable_kernel/include/ck_tile/ops/fmha
History
Po Yen Chen a2402950de [CK_TILE][FMHA] Add FP8 support for batch_prefill kernel (#3425) 
* Add fp8bf16 support for batch_prefill

* Fix wrong scale_s re-compute logic in batch_prefill

* Fix wrong scale_s re-compute logic in fmha fwd

* Fix batch_prefill codegen error

* Remove no-longer used GetName() function

* Add fp8 logits=True instances

* Update CHANGELOG.md

[ROCm/composable_kernel commit: 1c3151963b]
2025-12-24 10:34:06 +08:00
..
block
Add attention sink support for FMHA FWD (#3368)
2025-12-15 12:21:59 +08:00
kernel
[CK_TILE][FMHA] Add FP8 support for batch_prefill kernel (#3425)
2025-12-24 10:34:06 +08:00
pipeline
[CK_TILE][FMHA] Add logits soft-capping support for FAv3 (WIP) (#3355)
2025-12-18 16:08:45 +08:00