Kawrakow 6ef4954612 CUDA: better MoE implementation (#283) ... * Make fused MoE reproducible As a bonus, peak performance at pp2048 with u_batch = 2048 is ~8% better. * Slightly better * Also do it for non-fused mul_mat_id --------- Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>		2025-03-25 07:47:10 +01:00
..
cmake	Merge mainline llama.cpp (#3)	2024-07-27 07:55:01 +02:00
include	SER - Smart Expert Reduction (#239)	2025-03-02 13:47:38 +02:00
src	CUDA: better MoE implementation (#283)	2025-03-25 07:47:10 +01:00
.gitignore	Merge mainline llama.cpp (#3)	2024-07-27 07:55:01 +02:00
CMakeLists.txt	Compile time option to use bf16 for qunts without MMQ kernels (#261)	2025-03-18 07:37:10 +01:00