karpathy.py 文件源码

def __init__(self, env, monitor_path, video=True, **usercfg):
        super(Karpathy, self).__init__(**usercfg)
        self.env = wrappers.Monitor(env, monitor_path, force=True, video_callable=(None if video else False))
        self.nA = self.env.action_space.n
        # Default configuration. Can be overwritten using keyword arguments.
        self.config.update(dict(
            # timesteps_per_batch=10000,
            # n_iter=100,
            episode_max_length=env.spec.tags.get("wrapper_config.TimeLimit.max_episode_steps"),
            gamma=0.99,
            learning_rate=0.05,
            batch_size=10,  # Amount of episodes after which to adapt gradients
            decay_rate=0.99,  # Used for RMSProp
            n_hidden_units=20,
            draw_frequency=50,  # Draw a plot every 50 episodes
            repeat_n_actions=1
        ))
        self.config.update(usercfg)
        self.build_network()