from datasets import load_dataset
from transformers import AutoTokenizer

data_cnn = load_dataset("cnn_dailymail", '3.0.0')
data_xsum = load_dataset("xsum")

print(data_xsum)