You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by GitBox <gi...@apache.org> on 2019/08/13 03:48:23 UTC
[GitHub] [incubator-mxnet] ElaineBao commented on a change in pull request
#15870: update support MKLDNN BN conditions
ElaineBao commented on a change in pull request #15870: update support MKLDNN BN conditions
URL: https://github.com/apache/incubator-mxnet/pull/15870#discussion_r313210278
##########
File path: src/operator/nn/batch_norm.cc
##########
@@ -420,8 +419,7 @@ void BatchNormGradComputeExCPU(const nnvm::NodeAttrs &attrs,
mxnet::TShape shape = inputs[0].shape();
// MKLDNN batchnorm only works well on the special MKLDNN layout.
- if (SupportMKLDNNBN(inputs[0], param)
- && (inputs[3].IsMKLDNNData() || inputs[0].IsMKLDNNData())) {
+ if (SupportMKLDNNBN(inputs[0], param)) {
Review comment:
It's the same reason as line 399, as described in Performance part, when a model has fp32 bn along with other int8 ops, it triggers the IsMKLDNNData() check and cannot run int MKLDNN, which cause a slower speed.
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services