from datasets import load_dataset

data = load_dataset("nixiesearch/ms_marco")
print(data["train"].features)