From 4c72ab13b20c1589c3af7fc18d14299c983ad556 Mon Sep 17 00:00:00 2001
From: Georgi Gerganov <ggerganov@gmail.com>
Date: Thu, 28 Sep 2023 16:02:20 +0300
Subject: [PATCH] metal : use mm kernels for batch size > 2

---
 ggml-metal.m | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/ggml-metal.m b/ggml-metal.m
index 71d69d38d..b3c463f03 100644
--- a/ggml-metal.m
+++ b/ggml-metal.m
@@ -958,7 +958,7 @@ void ggml_metal_graph_compute(
                                 src1t == GGML_TYPE_F32 &&
                                 [ctx->device supportsFamily:MTLGPUFamilyApple7] &&
                                 ne00%32 == 0 &&
-                                ne11 > 1) {
+                                ne11 > 2) {
                                 switch (src0->type) {
                                     case GGML_TYPE_F32:  [encoder setComputePipelineState:ctx->pipeline_mul_mm_f32_f32];  break;
                                     case GGML_TYPE_F16:  [encoder setComputePipelineState:ctx->pipeline_mul_mm_f16_f32];  break;