From 70000b47e2d3ca2f581604e70c870156146a0eb1 Mon Sep 17 00:00:00 2001 From: Concedo <39025047+LostRuins@users.noreply.github.com> Date: Wed, 26 Jun 2024 16:47:12 +0800 Subject: [PATCH] Revert "Revert "set flags to optimize for mmq"" This reverts commit 7959e937a10bb935d5682701ae01e5a51c3157e3. --- ggml-cuda/mmq.cuh | 1 + 1 file changed, 1 insertion(+) diff --git a/ggml-cuda/mmq.cuh b/ggml-cuda/mmq.cuh index d59c0532f..f97bbd6e7 100644 --- a/ggml-cuda/mmq.cuh +++ b/ggml-cuda/mmq.cuh @@ -8,6 +8,7 @@ #include #define MMQ_DP4A_MAX_BATCH_SIZE 64 // Max. batch size to use for dp4a MMQ kernels when FP16 tensor cores are available. +#define GGML_CUDA_FORCE_MMQ typedef void (*load_tiles_mmq_t)(const char * __restrict__ x, int * x_tile, const int & kbx0, const int & i_max, const int & stride); typedef void (*vec_dot_mmq_t)(const int * __restrict__ x, const int * __restrict__ y, float * __restrict__ sum, const int & k0);