ggml : when using BLAS start only 1 CPU thread

diff --git a/ggml.c b/ggml.c
index 058241e7f9dcdba86b4e1845367a8485c5dc5587..f4c96eb42229d494905fb538fd9d0e3e3618a3f9 100644 (file)
--- a/ggml.c
+++ b/ggml.c
@@ -7257,6 +7257,7 @@ void ggml_graph_compute(struct ggml_context * ctx, struct ggml_cgraph * cgraph)
                                 node->src1->type == GGML_TYPE_F32) {
 #if defined(GGML_USE_ACCELERATE) || defined(GGML_USE_OPENBLAS)
                                 if (ggml_compute_forward_mul_mat_use_blas(node->src0, node->src1, node)) {
+                                    node->n_tasks = 1;
                                     cur = sizeof(float)*(node->src0->ne[0]*node->src0->ne[1]);
                                 } else {
                                     cur = sizeof(ggml_fp16_t)*ggml_nelements(node->src1);