Fix condition for broadcast

Author: am17an

ggml/src/ggml-cuda/ggml-cuda.cu

@@ -2786,8 +2786,7 @@ static bool ggml_cuda_can_fuse(const struct ggml_cgraph * cgraph, int node_idx,
         }
 
         //if rms norm is the B operand, then we don't handle broadcast
-        if (rms_norm == mul->src[1] &&
-            mul->src[0]->ne[1] != rms_norm->src[1]->ne[1]) {
+        if (rms_norm == mul->src[1] && !ggml_are_same_shape(mul->src[0], rms_norm->src[1])) {
             return false;
         }