def main():
agent = QLearningAgent()
if os.path.isfile(FILE):
with open(FILE, 'r') as f:
agent.Q = pickle.load(f)
for i in range(NUM_GAMES):
print(i)
agent.train()
with open(FILE, 'w') as f:
agent.save(f)
# time.sleep(0.5)
env.render()
if done:
break
return total_reward
# 使用gym创建迷宫环境，设置is_slippery为False降低环境难度
env = gym.make("FrozenLake-v0", is_slippery=False) # 0 left, 1 down, 2 right, 3 up
# 创建一个agent实例，输入超参数
agent = QLearningAgent(
obs_n=env.observation_space.n,
act_n=env.action_space.n,
learning_rate=0.1,
gamma=0.9,
e_greed=0.1)
# 训练500个episode，打印每个episode的分数
for episode in range(500):
ep_reward, ep_steps = run_episode(env, agent, True)
print('Episode %s: steps = %s , reward = %.1f' % (episode, ep_steps, ep_reward))
# 全部训练结束，查看算法效果
test_reward = test_episode(env, agent)
# 保存Q table
if(test_reward == 1):
agent.save()
print('test reward = %.1f' % (test_reward))