The BigScience ROOTS Corpus: A 1.6TB Composite Multilingual Dataset
Name
the_bigscience_roots_corpus_a_.pdf
Description
visibility:open
Size
2.27 MB
Format
Adobe PDF
Checksum (CRC64NVME)
OT5K3SA2j+Y=
Resource type
Conference paper (published)
Creator (person)
Laurenà§on, Hugo
Saulnier, Lucile
Wang, Thomas
Akiki, Christopher
Villanova del Moral, Albert
Le Scao, Teven
van Strien, Daniel
Date published
2022
Abstract
As language models grow ever larger, the need for large-scale high-quality text datasets has never been more pressing, especially in multilingual settings. The BigScience workshop, a 1-year international and multidisciplinary initiative, was formed with the goal of researching and training large language models as a values-driven undertaking, putting issues of ethics, harm, and governance in the foreground. This paper documents the data creation and curation efforts undertaken by BigScience to assemble the Responsible Open-science Open-collaboration Text Sources (ROOTS) corpus, a 1.6TB dataset spanning 59 languages that was used to train the 176-billion-parameter BigScience Large Open-science Open-access Multilingual (BLOOM) language model. We further release a large initial subset of the corpus and analyses thereof, and hope to empower large-scale monolingual and multilingual modeling projects with both the data and the processing tools, as well as stimulate research around this large multilingual corpus.
Project(s)
Living with Machines
Funder
| Funder name | Awards |
Arts and Humanities Research Council | AH/S01179X/1 |
Event title
Thirty-sixth Conference on Neural Information Processing Systems Datasets and Benchmarks Track
Publisher
Neural Information Processing Systems
Place of publication
USA
Rights statement
In Copyright
Keywords