kinoml.ml.tensorflow_models

Example models for TensorFlow

Note

This code is not currently in use.

Module Contents

kinoml.ml.tensorflow_models.DNN(input_dim)

DNN builds and compiles a TF model (a Deep Neural Network) that takes as input ‘input_dim’

Parameters

input_dim (tuple of int) – Expected shape of the input data

Returns

model

Return type

tf.keras.models.Sequential

kinoml.ml.tensorflow_models.CNN(input_shape)

CNN builds and compiles a TF model (a Convolutional Neural Network) that takes as input ‘input_shape’ :param input_shape: Expected shape of the input data :type input_shape: tuple of int

Returns

model

Return type

tf.keras.models.Sequential

kinoml.ml.tensorflow_models.MPNN(input_shape)

MPNN builds and compiles a TF model (a Message Passing Neural Network) that takes as input ‘input_shape’ :param input_shape: Expected shape of the input data :type input_shape: tuple of int

Returns

model

Return type

tf.keras.models.Sequential