zheng-da commented on a change in pull request #8302: Refactor operators & 
MKLDNN
URL: https://github.com/apache/incubator-mxnet/pull/8302#discussion_r164203240
 
 

 ##########
 File path: src/kvstore/kvstore_dist.h
 ##########
 @@ -32,11 +32,6 @@
 #include "mxnet/engine.h"
 #include "ps/ps.h"
 #include "./kvstore_dist_server.h"
-#if MKL_EXPERIMENTAL == 1
 
 Review comment:
   Thank you for your instructions. I run the test on a GPU instance and it 
seems to pass the test. It seems the test is specifically written for sparse 
formats and gradient compression.
   
   Currently, MKLDNN always uses the default layout for weight arrays during 
the training. I guess it may make more sense to train a model in a distributed 
environment?

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on 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

Reply via email to