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 2017/12/08 01:17:51 UTC

[GitHub] rahul003 commented on issue #8979: Question about 2bit gradient compression

rahul003 commented on issue #8979: Question about 2bit gradient compression  
URL: https://github.com/apache/incubator-mxnet/issues/8979#issuecomment-350143925
 
 
   Yes currently pull doesn't compress gradients to avoid compression at two steps. I suspect this would lead to poorer convergence because we would be losing more information. We can do this if experiments show that accuracy isn't affected much. 

----------------------------------------------------------------
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:
users@infra.apache.org


With regards,
Apache Git Services