Kawrakow 727f7b7d9f Refactor CUDA flash attention (#745) ... * Factor out mma * Factor out wmma * Factor out vec * Remove unnecessary includes from fattn.cu * Move mma launch to fattn-mma-f16.cuh * Slightly better PP --------- Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>		2025-09-02 10:12:56 +02:00
..
cmake	Merge mainline llama.cpp (#3)	2024-07-27 07:55:01 +02:00
include	Fused FFN_UP+FFN_GATE op (#741)	2025-08-31 18:16:36 +03:00
src	Refactor CUDA flash attention (#745)	2025-09-02 10:12:56 +02:00
.gitignore	Merge mainline llama.cpp (#3)	2024-07-27 07:55:01 +02:00
CMakeLists.txt	Set default value of GGML_SCHED_MAX_COPIES to 1 (#751)	2025-09-02 07:04:39 +02:00