Preprint
MLS: A Large-Scale Multilingual Dataset for Speech Research
Vineel Pratap et al., arXiv (Computation and Language), 2020
Read the originalRead on October 17, 2026
- What it is
- The paper that introduced Multilingual LibriSpeech, a corpus of read audiobooks in eight European languages, with about 44.5 thousand hours of English and about 6 thousand hours in the other languages.
- How much weight it carries
- A preprint from Facebook AI Research describing a public dataset that is widely used as a benchmark. The hours are approximate totals given in the abstract. The data is read audiobook speech, the same genre as LibriSpeech.
Facts taken from this source
- Multilingual LibriSpeech has about 44.5 thousand hours of English audio, drawn from read audiobooks, and about 6 thousand hours across its other languages.
“consists of 8 languages, including about 44.5K hours of English and a total of about 6K hours for other languages” (Abstract)