RL4J

gistfile1.txt

/**
 * Created by danlin on 2016-08-17.
 */
import org.deeplearning4j.rl4j.gym.space.Box;
import org.deeplearning4j.rl4j.learning.ILearning;
import org.deeplearning4j.rl4j.learning.sync.qlearning.QLearning;
import org.deeplearning4j.rl4j.learning.sync.qlearning.discrete.QLearningDiscreteDense;
import org.deeplearning4j.rl4j.mdp.gym.GymEnv;
import org.deeplearning4j.rl4j.network.dqn.DQNFactoryStdDense;
import org.deeplearning4j.rl4j.policy.Policy;
import org.deeplearning4j.rl4j.space.DiscreteSpace;
import org.deeplearning4j.rl4j.util.DataManager;


    public class Cartpole
    {
        public static String OPENAI_KEY = "";

        public static QLearning.QLConfiguration CARTPOLE_QL =
                new QLearning.QLConfiguration(
                        123,
                        500,
                        150000,
                        100000,
                        32,
                        100,
                        1000,
                        0.99,
                        100.0,
                        0.0f,
                        1f / 40000f,
                        true
                );

        public static DQNFactoryStdDense.Configuration CARTPOLE_NET =
                new DQNFactoryStdDense.Configuration(4, 0.001, 0.001,  0.99);

        public static void main( String[] args )
        {
            cartPole();

        }

        public static void cartPole() {
            DataManager manager = new DataManager(true);
            GymEnv mdp = new GymEnv("CartPole-v0", false);
            ILearning<Box, Integer, DiscreteSpace> dql = new QLearningDiscreteDense(mdp, CARTPOLE_NET, CARTPOLE_QL, manager);
            dql.train();
            Policy<Box, Integer> pol = dql.getPolicy();
            while(true){
                mdp.reset();
                pol.play(mdp);
            }

        }


    }