reminisce commented on issue #14998: Revert the change on broadcast_to param 
shape
URL: https://github.com/apache/incubator-mxnet/pull/14998#issuecomment-495459567
 
 
   @DickJC123 Thanks for the review. For this specific problem, it actually has 
nothing to do with numpy compatibility mode. The document says that `0` just 
means the output dimension size is the same as the corresponding input 
dimension size; it's not related to any judgement on whether the shape 
dimension size should be 0 or unknown. So per document, `0` should never be 
changed to `-1` in whatever situation, but it was changed by a mistake. This is 
the same as MXNet `reshape` operator where the shape parameter has `0`, `-1`, 
`-2`, `-3`, and `-4` for different shape inference maneuvers.

----------------------------------------------------------------
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:
[email protected]


With regards,
Apache Git Services

Reply via email to