experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Fix JSON serialization error in TensorFlowOpLayer in Python 3.

Browse Source 
PiperOrigin-RevId: 259397921
This commit is contained in:
Thomas O'Malley 2019-07-22 13:46:02 -07:00 committed by TensorFlower Gardener
parent 3c5fb53765
commit 7cc180f107
1 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
tensorflow/python/keras/engine/base_layer.py
View File

@ -2387,6 +2387,8 @@ class TensorFlowOpLayer(Layer):
dtype=None):
super(TensorFlowOpLayer, self).__init__(
name=_TF_OP_LAYER_NAME_PREFIX + name, trainable=trainable, dtype=dtype)
if not isinstance(node_def, bytes):
node_def = node_def.encode('utf-8')
self.node_def = node_def_pb2.NodeDef.FromString(node_def)
self.constants = constants or {}
# Layer uses original op unless it is called on new inputs.
@ -2446,7 +2448,7 @@ class TensorFlowOpLayer(Layer):
def get_config(self):
config = super(TensorFlowOpLayer, self).get_config()
config.update({
'node_def': self.node_def.SerializeToString(),
'node_def': self.node_def.SerializeToString().decode('utf-8'),
'constants': {
i: backend.get_value(c) for i, c in self.constants.items()
}