cudnn now provides LSTMP implementation in 7.2. https://docs.nvidia.com/deeplearning/sdk/cudnn-developer-guide/index.html#features-of-rnn-functions
Since this will require operator API changes, let's discuss in #9543. [ Full content available at: https://github.com/apache/incubator-mxnet/issues/11253 ] This message was relayed via gitbox.apache.org for [email protected]
