from datasets import load_dataset
eli5 = load_dataset("eli5_category", split="train[:5000]")

Split the dataset's train split into a train and test set with the [~datasets.Dataset.train_test_split] method:

eli5 = eli5.train_test_split(test_size=0.2)

Then take a look at an example:

eli5["train"][0]
{'q_id': '7h191n',
 'title': 'What does the tax bill that was passed today mean?