[incubator-mxnet] branch master updated: modifying SyncBN doc for FP16 use case (#14041)

[sr...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

srochel pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git


The following commit(s) were added to refs/heads/master by this push:
     new 3f6778b  modifying SyncBN doc for FP16 use case (#14041)
3f6778b is described below

commit 3f6778b0e88c88d010f40cb3555c822430496acb
Author: Manu Seth <22...@users.noreply.github.com>
AuthorDate: Mon Feb 4 20:53:05 2019 -0800

    modifying SyncBN doc for FP16 use case (#14041)
    
    LGTM
---
 python/mxnet/gluon/contrib/nn/basic_layers.py | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/python/mxnet/gluon/contrib/nn/basic_layers.py b/python/mxnet/gluon/contrib/nn/basic_layers.py
index 28fea15..56f0809 100644
--- a/python/mxnet/gluon/contrib/nn/basic_layers.py
+++ b/python/mxnet/gluon/contrib/nn/basic_layers.py
@@ -165,7 +165,10 @@ class SyncBatchNorm(BatchNorm):
 
     Standard BN [1]_ implementation only normalize the data within each device.
     SyncBN normalizes the input within the whole mini-batch.
-    We follow the sync-onece implmentation described in the paper [2]_.
+    We follow the implementation described in the paper [2]_.
+
+    Note: Current implementation of SyncBN does not support FP16 training.
+    For FP16 inference, use standard nn.BatchNorm instead of SyncBN.
 
     Parameters
     ----------