metal : do not use mul_mm kernels when ne00 < 64 (#3542)

diff --git a/ggml-metal.m b/ggml-metal.m
index 29cb3c922daeba4e912f1b2defcca6fbc30e18f4..7d67db90f078b20782d490d0361196163976e457 100644 (file)
--- a/ggml-metal.m
+++ b/ggml-metal.m
@@ -1040,7 +1040,7 @@ void ggml_metal_graph_compute(
                                 !ggml_is_transposed(src0) &&
                                 !ggml_is_transposed(src1) &&
                                 src1t == GGML_TYPE_F32 &&
-                                ne00 % 32 == 0 &&
+                                ne00 % 32 == 0 && ne00 >= 64 &&
                                 ne11 > ne11_mm_min) {
                                 //printf("matrix: ne00 = %6d, ne01 = %6d, ne02 = %6d, ne11 = %6d, ne12 = %6d\n", ne00, ne01, ne02, ne11, ne12);
                                 switch (src0->type) {