| 70 | self.env_id = self.remotes[0].recv() |
| 71 | |
| 72 | def step(self, actions): |
| 73 | for remote, action in zip(self.remotes, actions): |
| 74 | remote.send(('step', action)) |
| 75 | results = [remote.recv() for remote in self.remotes] |
| 76 | obs, rews, dones, infos = zip(*results) |
| 77 | # print("Infos:", infos) |
| 78 | # for done, info in zip(dones, infos): |
| 79 | # if done: |
| 80 | # # print("Total reward:", info['reward'], "Num steps:", info['episode_length']) |
| 81 | # print("Returned info:", info, "Done:", done) |
| 82 | return np.stack(obs), np.stack(rews), np.stack(dones), infos |
| 83 | |
| 84 | def reset(self): |
| 85 | for remote in self.remotes: |