TaoLv commented on a change in pull request #14545: Optimize transpose operator with MKL-DNN URL: https://github.com/apache/incubator-mxnet/pull/14545#discussion_r270672007
########## File path: src/operator/nn/mkldnn/mkldnn_transpose.cc ########## @@ -0,0 +1,156 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See the NOTICE file + * distributed with this work for additional information + * regarding copyright ownership. The ASF licenses this file + * to you under the Apache License, Version 2.0 (the + * "License"); you may not use this file except in compliance + * with the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, + * software distributed under the License is distributed on an + * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY + * KIND, either express or implied. See the License for the + * specific language governing permissions and limitations + * under the License. + */ + +/*! + * \file mkldnn_transpose.cc + * \brief + * \author +*/ + +#if MXNET_USE_MKLDNN == 1 + +#include <mkldnn.hpp> +#include "../../tensor/matrix_op-inl.h" + +namespace mxnet { +namespace op { + +bool SupportMKLDNNTranspose(const TransposeParam& param, + const NDArray &data) { + auto data_ndim = data.shape().ndim(); + + if (data_ndim > 4 || data.dtype() != mshadow::kFloat32) Review comment: It should work but it's not tested and verified. So here I limited the dimensionality and data type just like what we did for other MKL-DNN operators. BTW, if we want to use INT8 transpose in a quantized network, probably we need a quantized transpose operator to accept and output an additional scale argument. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
