641a28fbf0daff0ad1ad0f43d2c4b545cb6f9656,examples/reinforcement_learning/tutorial_cartpole_ac.py,Actor,choose_action,#Actor#,110

Before Change

        return exp_v

    def choose_action(self, s):
        probs = self.sess.run(self.acts_prob, {self.s: [s]})  // get probabilities of all actions
        return tl.rein.choice_action_by_probs(probs.ravel())

    def choose_action_greedy(self, s):

After Change

    def choose_action(self, s):
            // probs = self.sess.run(self.acts_prob, {self.s: [s]})  // get probabilities of all actions
        _logits = self.model([s]).outputs
        _probs = tf.nn.softmax(_logits).numpy()
        return tl.rein.choice_action_by_probs(_probs.ravel())

    def choose_action_greedy(self, s):

Instances