[GitHub] austingg commented on issue #7393: add depthwise convolution's gpu version optimization

2017-08-29 Thread git 
austingg commented on issue #7393: add depthwise convolution's gpu version 
optimization
URL: https://github.com/apache/incubator-mxnet/pull/7393#issuecomment-325631062
 
 
   @terrychenism  how about the convergence, I used the symbol in 
`imageclassification` dir and script `train_imagenet.py`, seems not converge 
when using depthwise conv.
 

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:
us...@infra.apache.org


With regards,
Apache Git Services

[GitHub] austingg commented on issue #7393: add depthwise convolution's gpu version optimization

2017-08-09 Thread git 
austingg commented on issue #7393: add depthwise convolution's gpu version 
optimization
URL: https://github.com/apache/incubator-mxnet/pull/7393#issuecomment-321261481
 
 
   @solin319 , it is said grouped conv in cudnn v7 is not efficient when used 
in depthwise conv, where the group num == input channels
 

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:
us...@infra.apache.org


With regards,
Apache Git Services