abb4e425ff57acad92d42e92a6c76526a3b49e27,softlearning/samplers/simple_sampler.py,SimpleSampler,sample,#SimpleSampler#,19

Before Change

            next_observations=next_observation)

        if terminal or self._path_length >= self._max_path_length:
            last_path = self.pool.last_n_batch(
                self._path_length,
                observation_keys=getattr(self.env, "observation_keys", None))
            last_path.update({"infos": self._infos})
            self._last_n_paths.appendleft(last_path)

            self.policy.reset()

After Change

        self._current_path["observations"].append(self._current_observation)
        self._current_path["actions"].append(action)
        self._current_path["rewards"].append([reward])
        self._current_path["terminals"].append([terminal])
        self._current_path["next_observations"].append(next_observation)
        self._current_path["infos"].append(info)

Instances