ray.data.DatasetPipeline.take_all#

DatasetPipeline.take_all(limit: Optional[int] = None) → List[ray.data.block.T][source]#

Call Dataset.take_all over the stream of output batches from the pipeline