[graph] Skip backwarding bug fix
authorParichay Kapoor <pk.kapoor@samsung.com>
Mon, 1 Nov 2021 06:51:42 +0000 (15:51 +0900)
committerJijoong Moon <jijoong.moon@samsung.com>
Wed, 3 Nov 2021 06:50:53 +0000 (15:50 +0900)
commitd1b9befdbb0d15d6681dd4707f42b64ebb998176
tree63051556c35b177db036a2fe7e850ca3e71be263
parentc93c4712d446a74b8794cdc33ace0668fc4621bb
[graph] Skip backwarding bug fix

Current implementation assumes that the graph is always linear and skips
the backwarding for contiguous set of layers at the start of the model
which should not do the backwarding.

This patch makes a fix to this issue but allowing layers in the middle
of the sorted graph to skip backwarding.

Signed-off-by: Parichay Kapoor <pk.kapoor@samsung.com>
nntrainer/graph/network_graph.cpp
nntrainer/graph/network_graph.h
nntrainer/layers/layer_node.cpp
nntrainer/layers/layer_node.h
nntrainer/models/neuralnet.cpp
test/unittest/models/models_test_utils.cpp
test/unittest/models/models_test_utils.h