(self, policy)
| 131 | self.full = True |
| 132 | |
| 133 | def update_values(self, policy): |
| 134 | for i in range(self.buffer_size): |
| 135 | obs_dict = {} |
| 136 | for k in self.observations.keys(): |
| 137 | obs_dict[k] = self.observations[k][i] |
| 138 | values = policy.forward_value(obs_dict) |
| 139 | self.values[i] = values |
| 140 | |
| 141 | def get(self, batch_size: Optional[int] = None) -> Generator[PpoBufferSamples, None, None]: |
| 142 | assert self.full, '' |