Xingjian Shi created MXNET-58:
---------------------------------
Summary: Add LayerNorm in MXNet
Key: MXNET-58
URL: https://issues.apache.org/jira/browse/MXNET-58
Project: Apache MXNet
Issue Type: New Feature
Reporter: Xingjian Shi
# Directly implement layer normalization in C++. The speed and memory cost are
both better than the way of stacking the broadcast/reduce OPs. Solves
[#9950|https://github.com/apache/incubator-mxnet/issues/9950]
# Add LayerNorm in Gluon
# Fix the doc of InstanceNorm. In InstanceNorm, the real axis to normalize the
input tensor is all axes excluding the 0th axis and the given axis.
# Fix the doc of BatchNorm, the inverse std instead of the var is set as the
output. Should fix [#9216|https://github.com/apache/incubator-mxnet/issues/9216]
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
