(self, state)
| 280 | |
| 281 | |
| 282 | def act(self, state): |
| 283 | if np.random.rand() <= self.epsilon: |
| 284 | return np.random.choice(self.action_size) |
| 285 | act_values = self.model.predict(state, verbose=0) |
| 286 | return np.argmax(act_values[0]) # returns action |
| 287 | |
| 288 | |
| 289 | def replay(self, batch_size=32): |
no test coverage detected