wetliu commented on issue #7064: is "num_group" implement mobilenet?
   I tested the speed between TF depth wise convolution with the 1*1 
convolution and the normal convolution. The speed of depth wise conv with the 
1*1 operator is 2 times as fast as the normal convolution. I think the group 
trick is similar to the depth-wise convolution, but it seems there is some 
difference (I am not very sure).
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:

With regards,
Apache Git Services

Reply via email to