x110 commented on issue #12899: Gradient of BatchNorm layer URL: https://github.com/apache/incubator-mxnet/issues/12899#issuecomment-431900201 For anyone else wondering `fix_gamma=True` is used when the next layer is linear (also relu) because then the scaling can be done by that layer.
---------------------------------------------------------------- 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
