reinforce.py 文件源码

def build_network_normal(self):
        # Symbolic variables for observation, action, and advantage
        self.states = tf.placeholder(tf.float32, [None, self.env_runner.nO], name="states")  # Observation
        self.a_n = tf.placeholder(tf.float32, name="a_n")  # Continuous action
        self.adv_n = tf.placeholder(tf.float32, name="adv_n")  # Advantage

        L1 = tf.contrib.layers.fully_connected(
            inputs=self.states,
            num_outputs=self.config["n_hidden_units"],
            activation_fn=tf.tanh,
            weights_initializer=tf.truncated_normal_initializer(mean=0.0, stddev=0.02),
            biases_initializer=tf.zeros_initializer())

        mu, sigma = mu_sigma_layer(L1, 1)

        self.normal_dist = tf.contrib.distributions.Normal(mu, sigma)
        self.action = self.normal_dist.sample(1)
        self.action = tf.clip_by_value(self.action, self.env.action_space.low[0], self.env.action_space.high[0])