39784035924760e69ec50ab6ab29c8666a595eec,softlearning/replay_pools/replay_pool.py,ReplayPool,add_path,#ReplayPool#Any#,34

Before Change

        :param path: Dict like one outputted by railrl.samplers.util.rollout
        
        for i, (
                observation,
                action,
                reward,
                next_observation,
                terminal,
                agent_info,
                env_info
        ) in enumerate(zip(
            path["observations"],
            path["actions"],
            path["rewards"],
            path["next_observations"],
            path["terminals"],
            path.get("agent_infos", {}),
            path.get("env_infos", {}),
        )):
            self.add_sample(
                observation=observation,
                action=action,
                reward=reward,
                next_observation=next_observation,
                terminal=terminal,
                agent_info=agent_info,
                env_info=env_info,
            )
        self.terminate_episode()

    @abc.abstractmethod
    def random_batch(self, batch_size):

After Change

        path_length = path["observations"].shape[0]
        self.add_samples(num_samples=path_length, **{
            key: value
            for key, value in path.items()
            if key in self.field_names
        })
        self.terminate_episode()

Instances