Merge pull request #595 from lilleswing/flaky

               use_queue=True,

               mode="regression",

               graph=None,

               learning_rate=0.001,

               **kwargs):

    """

    TODO(LESWING) allow a model to change its learning rate

    self.built = False

    self.queue_installed = False

    self.optimizer = TFWrapper(

        tf.train.AdamOptimizer, learning_rate=0.001, beta1=0.9, beta2=0.999)

        tf.train.AdamOptimizer,

        learning_rate=learning_rate,

        beta1=0.9,

        beta2=0.999)

    # Singular place to hold Tensor objects which don't serialize

    # These have to be reconstructed on restoring from pickle

from tensorflow.python.framework import test_util

from sklearn.ensemble import RandomForestClassifier

from sklearn.ensemble import RandomForestRegressor

from flaky import flaky

class TestOverfit(test_util.TensorFlowTestCase):

    scores = model.evaluate(dataset, [classification_metric])

    assert scores[classification_metric.name] > .9

  @flaky

  def test_tg_multitask_classification_overfit(self):

    """Test TensorGraph multitask overfits tiny data."""

    n_tasks = 10

    total_loss = ReduceMean(in_layers=entropies)

    tg = dc.models.TensorGraph(learning_rate=0.01)

    tg = dc.models.TensorGraph(learning_rate=0.01, batch_size=n_data_points)

    for output in outputs:

      tg.add_output(output)

    tg.set_loss(total_loss)