(self, state, sess)
| 169 | return |
| 170 | |
| 171 | def sample_action(self, state, sess): |
| 172 | # Make input N x D (N = 1) |
| 173 | feed_dict = { self.policy_net.states: [state] } |
| 174 | actions = sess.run(self.policy_net.sample_action, feed_dict) |
| 175 | # Prediction is a 1-D array of length N, just want the first value |
| 176 | return actions[0] |
| 177 | |
| 178 | def get_value_prediction(self, state, sess): |
| 179 | # Make input N x D (N = 1) |