projects / platform / core / ml / nntrainer.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 84c6611) | patch
[batchnorm] Optimize batch normalization implementation
authorParichay Kapoor <pk.kapoor@samsung.com>
Fri, 24 Sep 2021 09:08:38 +0000 (18:08 +0900)
committerJijoong Moon <jijoong.moon@samsung.com>
Fri, 1 Oct 2021 04:28:07 +0000 (13:28 +0900)
commit58d10086b11589854efc9a146bd0630473745d1c
tree4490070b8d8635a3abb29fb381f3d87612fd3516tree | snapshot
parent84c6611037832d4441cdbd2baa15a358c848becacommit | diff
[batchnorm] Optimize batch normalization implementation

This patch optimizes batch normalization implementation.
Reduces the temporary memory allocation and provide speedup for the
layer execution. With this patch, resnet18 runtime improves by approx
5%.

Signed-off-by: Parichay Kapoor <pk.kapoor@samsung.com>
nntrainer/layers/bn_layer.cpp diff | blob | history
Domain: Machine Learning / ML Framework;