Marco de CA de aprendizaje por refuerzo

import gym
import tensorflow as tf
import numpy as np
import random
from collections import deque

# Hyper Parameters
GAMMA = 0.95 # discount factor
LEARNING_RATE=0.01

class Actor():# PI
    def __init__(self, env, sess):
        # init some parameters
        self.time_step = 0
        self.state_dim = env.observation_space.shape[0]
        self.action_dim = env.action_space.n
        # 策略
        self.create_softmax_network()

        # Init session
        self.session = sess
        self.session.run(tf.global_variables_initializer())

    def create_softmax_network(self):
        # network weights
        W1 = self.weight_variable([self.state_dim, 20])
        b1 = self.bias_variable([20])
        W2 = self.weight_variable([20, self.action_dim])
        b2 = self.bias_variable([self.action_dim])
        # input layer
        self.state_input = tf.placeholder("float", [None, self.state_dim])
        self.tf_acts =