from datasets import load_dataset

data = load_dataset("nixiesearch/MSMARCO", split="train")