a3c.py 文件源码

def __init__(self,
            graph,
            exploration_steps,
            total_steps,
            gamma,
            a3c_update_interval,
            action_sampler):
        """
        graph should have the placeholders called "states", "actions",
        and "returns". It should also have operations called "loss_op", "train_op",
        "probs", and "value".
        """

        self.graph = graph
        self.gamma = gamma
        self.a3c_update_interval = a3c_update_interval
        self.action_sampler = action_sampler

        self.T = graph.get_collection("global_step")[0]
        self.exploration_steps = exploration_steps
        self.total_steps = total_steps
        self.incr_T = tf.assign_add(self.T, 1)