ray.data.DatasetPipeline.take#

DatasetPipeline.take(limit: int = 20) → List[ray.data.block.T][source]#

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