d5aba73e0c518742ad38a2482f7a3a567b0e406a,research/a3c_blogpost/a3c_cartpole.py,Worker,run,#Worker#,285

Before Change

      time_count = 0
      done = False
      while not done:
        logits, _ = self.local_model(tf_wrap(current_state[None, :]))
        probs = tf.nn.softmax(logits)

        action = np.random.choice(self.action_size, p=probs.numpy()[0])

After Change

      done = False
      while not done:
        logits, _ = self.local_model(
            tf.convert_to_tensor(current_state[None, :],
                                 dtype=tf.float32))
        probs = tf.nn.softmax(logits)

        action = np.random.choice(self.action_size, p=probs.numpy()[0])

Instances