Datasets:

Multilinguality:
multilingual
Size Categories:
1M<n<10M
Language Creators:
found
Annotations Creators:
found
Source Datasets:
original
ArXiv:
License:

Make the dataset streamable

#6
by lhoestq HF staff - opened

This way people can do

from datasets import load_dataset

ds = load_dataset("csebuetnlp/xlsum", "french", streaming=True)
for example in ds["train"]:
    ...

without downloading the full dataset

Ready to merge
This branch is ready to get merged automatically.

Sign up or log in to comment