Replace inputs.shape
with ops.shape(inputs)
in the layers
#18836
Labels
stat:awaiting keras-eng
Awaiting response from Keras engineer
type:feature
The user is asking for a new feature.
ops.shape(inputs)
is more stable especially for the TensorFlow backend.Some ops may lose the shape of the output tensor and contains a lot of
None
s in the shape.ops.shape(inputs)
can get the actual value of those dims.The text was updated successfully, but these errors were encountered: