Why why is this new order of operations for CUDA and cuDNN?

[ Full content available at: 
https://github.com/apache/incubator-mxnet/pull/12696 ]
This message was relayed via gitbox.apache.org for [email protected]

Reply via email to