cuda : use CUBLAS_COMPUTE_16F for non-attention ops

ggerganov · ggerganov · commit 0f2498f25d7e · 2023-10-27T20:19:42.000+03:00
diff --git a/ggml-cuda.cu b/ggml-cuda.cu
@@ -6385,19 +6385,27 @@ inline void ggml_cuda_op_mul_mat_cublas(
         }
         const half * src1_ptr = src1->type == GGML_TYPE_F16 ? (const half *) src1_ddq_i : src1_as_f16;
 
-        const float alpha = 1.0f;
-        const float beta = 0.0f;
+        size_t dst_as = 0;
+        half * dst_f16 = (half *) ggml_cuda_pool_malloc(row_diff*src1_ncols * sizeof(half), &dst_as);
+
+        const half alpha = 1.0f;
+        const half beta = 0.0f;
 
         CUBLAS_CHECK(cublasSetStream(g_cublas_handles[id], stream));
         CUBLAS_CHECK(
             cublasGemmEx(g_cublas_handles[id], CUBLAS_OP_T, CUBLAS_OP_N,
                     row_diff, src1_ncols, ne10,
                     &alpha, src0_ptr, CUDA_R_16F, ne00,
                             src1_ptr, CUDA_R_16F, ne10,
-                    &beta,  dst_dd_i, CUDA_R_32F, ldc,
-                    CUBLAS_COMPUTE_32F,
+                    &beta,   dst_f16, CUDA_R_16F, ldc,
+                    CUBLAS_COMPUTE_16F,
                     CUBLAS_GEMM_DEFAULT_TENSOR_OP));
 
+        const to_fp32_cuda_t to_fp32_cuda = ggml_get_to_fp32_cuda(GGML_TYPE_F16);
+        to_fp32_cuda(dst_f16, dst_dd_i, row_diff*src1_ncols, stream);
+
+        ggml_cuda_pool_free(dst_f16, dst_as);
+
         if (src0_as != 0) {
             ggml_cuda_pool_free(src0_as_f16, src0_as);
         }
