fixed example for MolGANEncoderLayer

  --------

  See: MolGANMultiConvolutionLayer for using in layers.

  vertices = 9

  nodes = 5

  edges = 5

  units = 128

  layer = MolGANConvolutionLayer(units=units,edges=edges)

  adjacency_tensor= layers.Input(shape=(vertices, vertices, edges))

  node_tensor = layers.Input(shape=(vertices,nodes))

  hidden1 = layer([adjacency_tensor,node_tensor])

  output = layer(hidden1)

  model = keras.Model(inputs=[adjacency_tensor,node_tensor], outputs=[output])

  >>> vertices = 9

  >>> nodes = 5

  >>> edges = 5

  >>> units = 128

  >>> layer = MolGANConvolutionLayer(units=units,edges=edges)

  >>> adjacency_tensor= layers.Input(shape=(vertices, vertices, edges))

  >>> node_tensor = layers.Input(shape=(vertices,nodes))

  >>> hidden1 = layer([adjacency_tensor,node_tensor])

  >>> output = layer(hidden1)

  >>> model = keras.Model(inputs=[adjacency_tensor,node_tensor], outputs=[output])

  References

  ----------

  Example

  --------

  >>> def create_discriminator(adjacency_tensor, node_tensor):

  vertices = 9

  edges = 5

  dropout_rate = .0

  dense = layers.Dropout(dropout_rate)(dense)

  output = layers.Dense(units=1)(dense)

        return keras.Model(inputs=[adjacency_tensor,node_tensor], outputs=[output])

  keras.Model(inputs=[adjacency_tensor,node_tensor], outputs=[output])

  References

  ----------