Merge pull request #634 from lilleswing/a3c-robust

from flaky import flaky

import deepchem as dc

from deepchem.models.tensorgraph.layers import Reshape, Variable, SoftMax

import numpy as np

class TestA3C(unittest.TestCase):

  @flaky

  def test_roulette(self):

    """Test training a policy for the roulette environment."""

        TestPolicy(),

        max_rollout_length=50,

        optimizer=dc.models.tensorgraph.TFWrapper(

            tf.train.AdamOptimizer, learning_rate=0.005))

            tf.train.AdamOptimizer, learning_rate=0.0001))

    a3c.fit(100000)

    # It should have learned that the expected value is very close to zero, and that the best

conda install -y joblib

yes | pip install six

conda install -y -c omnia mdtraj

conda install -y scikit-learn

conda install -y scikit-learn=0.18.1

conda install -y setuptools

conda install -y -c conda-forge keras=1.2.2

yes | pip install --upgrade $protobuf_url