projects / platform / core / ml / nntrainer.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: ed175d5) | patch
[optimzer/layers] Gradient dimension should match weight dim
authorParichay Kapoor <pk.kapoor@samsung.com>
Thu, 16 Jul 2020 11:39:29 +0000 (20:39 +0900)
committerJijoong Moon <jijoong.moon@samsung.com>
Thu, 16 Jul 2020 12:12:16 +0000 (21:12 +0900)
commit7b39bd4309450dbefb92667a1465783d2c6ba528
tree365dd8e2eb987dd4f250ef8b9fb8c50a521bd2d6tree | snapshot
parented175d5cce79f133e551c49b0722151df5765d09commit | diff
[optimzer/layers] Gradient dimension should match weight dim

Gradient dimension should match weight dimension
Currently optimizer applies averaging of gradients, which is not correct
Apply averaging of gradients before calling applyGradients

Resolves #280

Signed-off-by: Parichay Kapoor <pk.kapoor@samsung.com>
nntrainer/src/bn_layer.cpp diff | blob | history
nntrainer/src/conv2d_layer.cpp diff | blob | history
nntrainer/src/fc_layer.cpp diff | blob | history
nntrainer/src/optimizer.cpp diff | blob | history
Domain: Machine Learning / ML Framework;