good ish

import json

import time

import copy

import shutil

from deepchem.models.tensorgraph.layers import Flatten, Dense, SoftMax, \

    Variable, \

        return {'action_prob': probs, 'value': value}

def eval_tic_tac_toe(value_weight, games=10 ** 4, rollouts=10 ** 5):

def eval_tic_tac_toe(value_weight, num_epoch_rounds=1, games=10 ** 4, rollouts=10 ** 5):

    """

    Returns the average reward over 1k games after 10k rollouts

    :param value_weight:

    """

    env = TicTacToeEnvironment()

    policy = TicTacToePolicy()

    model_dir = "/tmp/tictactoe"

    shutil.rmtree(model_dir)

    avg_rewards = []

    for j in range(10):

        a3c = dc.rl.A3C(env, policy, entropy_weight=0.01, value_weight=value_weight, model_dir="/tmp/tictactoe")

    for j in range(num_epoch_rounds):

        a3c = dc.rl.A3C(env, policy, entropy_weight=0.01, value_weight=value_weight, model_dir=model_dir)

        a3c.optimizer = dc.models.tensorgraph.TFWrapper(tf.train.AdamOptimizer, learning_rate=0.01)

        try:

            a3c.restore()

def main():

    value_weight = 0.70

    score = eval_tic_tac_toe(value_weight)

    with open('tictactoe_converge_lambda1.json', 'w') as fout:

        fout.write(json.dumps(score))

    value_weight = 6.0

    score = eval_tic_tac_toe(value_weight, num_epoch_rounds=10)

    print(score)

if __name__ == "__main__":