bigscience-data/

roots_id_indo4b_jw300

ROOTS Subset: roots_id_indo4b_jw300

Dataset uid: indo4b_jw300

Indo4B consists of around 4B words, with around 250M sentences. The dataset covers both formal and colloquial Indonesian sentences compiled from 12 corpus, of which two corpus cover Indonesian colloquial language, eight corpus cover formal Indonesian language, and the rest have a mixed style, both colloquial and formal.\n

https://www.indobenchmark.com/… See the full description on the dataset page: https://huggingface.co/datasets/bigscience-data/roots_id_indo4b_jw300.