gan_model_inputs.py

def gan_model_inputs(real_dim, z_dim):
    """
    Creates the inputs for the model.
    
    Arguments:
    ----------
    :param real_dim: tuple containing width, height and channels
    :param z_dim: The dimension of Z
    ----------
    Returns:
    Tuple of (tensor of real input images, tensor of z (noise) data, Generator learning rate, Discriminator learning rate)
    """
    real_inputs = tf.placeholder(tf.float32, (None, *real_dim), name='real_inputs')
    z_inputs = tf.placeholder(tf.float32, (None, z_dim), name="z_inputs")
    generator_learning_rate = tf.placeholder(tf.float32, name="generator_learning_rate")
    discriminator_learning_rate = tf.placeholder(tf.float32, name="discriminator_learning_rate")
    
    return real_inputs, z_inputs, generator_learning_rate, discriminator_learning_rate