Maintaining Layer Weights in a Siamese network

Question

I am looking to create a Siamese network to analyze kinematic data:

#Variables for the for func
input_shapes = [[(None, 3), (None, 3), (None, 9)],