Changes'

        ], [y_b], [w_b])

class GraphConvKerasModel(tf.keras.Model):

class _GraphConvKerasModel(tf.keras.Model):

  def __init__(self,

               n_tasks,

    All arguments have the same meaning as in GraphConvModel.

    """

    super(GraphConvKerasModel, self).__init__()

    super(_GraphConvKerasModel, self).__init__()

    if mode not in ['classification', 'regression']:

      raise ValueError("mode must be either 'classification' or 'regression'")

               **kwargs):

    """The wrapper class for graph convolutions.

    Note that since the underlying GraphConvKerasModel class is

    Note that since the underlying _GraphConvKerasModel class is

    specified using imperative subclassing style, this model

    cannout make predictions for arbitrary outputs. 

    self.n_classes = n_classes

    self.batch_size = batch_size

    self.uncertainty = uncertainty

    model = GraphConvKerasModel(

    model = _GraphConvKerasModel(

        n_tasks,

        graph_conv_layers=graph_conv_layers,

        dense_layer_size=dense_layer_size,

    model = GraphConvModel(

        len(tasks), batch_size=batch_size, mode='classification')

    model.fit(dataset, nb_epoch=10)

    model.fit(dataset, nb_epoch=50)

    scores = model.evaluate(dataset, [metric], transformers)

    assert scores['mean-roc_auc_score'] >= 0.9

    batch_size = 50

    model = GraphConvModel(len(tasks), batch_size=batch_size, mode='regression')

    model.fit(dataset, nb_epoch=100)

    model.fit(dataset, nb_epoch=800)

    scores = model.evaluate(dataset, [metric], transformers)

    assert all(s < 0.1 for s in scores['mean_absolute_error'])

        dropout=0.1,

        uncertainty=True)

    model.fit(dataset, nb_epoch=100)

    model.fit(dataset, nb_epoch=500)

    # Predict the output and uncertainty.

    pred, std = model.predict_uncertainty(dataset)