ray.rllib.utils.replay_buffers.multi_agent_prioritized_replay_buffer.MultiAgentPrioritizedReplayBuffer.add

MultiAgentPrioritizedReplayBuffer.add(batch: SampleBatch | MultiAgentBatch | Dict[str, Any], **kwargs) → None

Adds a batch to the appropriate policy’s replay buffer.

Turns the batch into a MultiAgentBatch of the DEFAULT_POLICY_ID if it is not a MultiAgentBatch. Subsequently, adds the individual policy batches to the storage.

Parameters:

• batch – The batch to be added.

• **kwargs – Forward compatibility kwargs.

DeveloperAPI: This API may change across minor Ray releases.