ray.rllib.core.learner.learner_group.LearnerGroup.async_update#

LearnerGroup.async_update(batch: ray.rllib.policy.sample_batch.MultiAgentBatch, *, minibatch_size: Optional[int] = None, num_iters: int = 1, reduce_fn: Optional[Callable[[List[Mapping[str, Any]]], dict]] = <function _reduce_mean_results>) → Union[List[Mapping[str, Any]], List[List[Mapping[str, Any]]]][source]#

Asnychronously do gradient based updates to the Learner(s) with batch.

Parameters

  • batch – The data batch to use for the update.

  • minibatch_size – The minibatch size to use for the update.

  • num_iters – The number of complete passes over all the sub-batches in the input multi-agent batch.

  • reduce_fn – An optional callable to reduce the results from a list of the Learner actors into a single result. This can be any arbitrary function that takes a list of dictionaries and returns a single dictionary. For example you can either take an average (default) or concatenate the results (for example for metrics) or be more selective about you want to report back to the algorithm’s training_step. If None is passed, the results will not get reduced.

Returns

A list of list of dictionaries of results, where the outer list corresponds to separate calls to async_update, and the inner list corresponds to the results from each Learner(s). Or if the results are reduced, a list of dictionaries of the reduced results from each call to async_update that is ready.