DQN

dqn.py

# -*- coding: utf-8 -*-
import random
import gym
import numpy as np
from collections import deque
from keras.models import Sequential
from keras.layers import Dense
from keras.optimizers import RMSprop

EPISODES = 5000


class DQNAgent:
    def __init__(self, state_size, action_size):
        self.state_size = state_size
        self.action_size = action_size
        self.memory = deque(maxlen=10000)
        self.gamma = 0.9    # discount rate
        self.epsilon = 1.0  # exploration rate
        self.e_decay = .996
        self.e_min = 0.05
        self.learning_rate = 0.0001
        self.model = self._build_model()

    def _build_model(self):
        # Neural Net for Deep-Q learning Model
        model = Sequential()
        model.add(Dense(64, input_dim=self.state_size, activation='tanh'))
        model.add(Dense(128, activation='tanh', init='uniform'))
        model.add(Dense(128, activation='tanh', init='uniform'))
        model.add(Dense(self.action_size, activation='linear'))
        model.compile(loss='mse',
                      optimizer=RMSprop(lr=self.learning_rate))
        return model

    def remember(self, state, action, reward, next_state, done):
        self.memory.append((state, action, reward, next_state, done))

    def act(self, state):
        if np.random.rand() <= self.epsilon:
            return random.randrange(self.action_size)
        act_values = self.model.predict(state)
        print(act_values)
        return np.argmax(act_values[0])  # returns action

    def replay(self, batch_size):
        minibatch = random.sample(self.memory, batch_size)
        for state, action, reward, next_state, done in minibatch:
            target = self.model.predict(state)
            if done:
                target[0][action] = reward
            else:
                target[0][action] = reward + self.gamma * \
                            np.max(self.model.predict(next_state)[0])
            self.model.fit(state, target, nb_epoch=1, verbose=0)
        if self.epsilon > self.e_min:
            self.epsilon *= self.e_decay

if __name__ == "__main__":
    env = gym.make('CartPole-v0')
    state_size = env.observation_space.shape[0]
    action_size = env.action_space.n
    agent = DQNAgent(state_size, action_size)

    for e in range(EPISODES):
        state = env.reset()
        state = np.reshape(state, [1, state_size])
        for time in range(5000):
            env.render()
            action = agent.act(state)
            next_state, reward, done, _ = env.step(action)
            next_state = np.reshape(next_state, [1, state_size])
            reward = -100 if done else reward
            agent.remember(state, action, reward, next_state, done)
            state = next_state
            if done or time == 4999:
                print("episode: {}/{}, score: {}, e: {:.2}"
                        .format(e, EPISODES, time, agent.epsilon))
                break
        agent.replay(16)