- class ray.rllib.utils.replay_buffers.prioritized_replay_buffer.PrioritizedReplayBuffer(capacity: int = 10000, storage_unit: str = 'timesteps', alpha: float = 1.0, **kwargs)#
This buffer implements Prioritized Experience Replay.
The algorithm has been described by Tom Schaul et. al. in “Prioritized Experience Replay”. See https://arxiv.org/pdf/1511.05952.pdf for the full paper.
DeveloperAPI: This API may change across minor Ray releases.
Initializes a PrioritizedReplayBuffer instance.
Adds a batch of experiences or other data to this buffer.
Calls the given function with this Actor instance.
Returns the computer's network name.
Returns all local state.
Ping the actor.
num_itemsitems from this buffer, including prio.
Restores all local state to the provided
Returns the stats of this buffer.
Update priorities of items at given indices.