opus_ubuntu
102A parallel corpus of Ubuntu localization files. Source: https://translations.launchpad.net 244 languages, 23,988 bitexts total number of files: 30,959 total number of tokens:
Datasets
3
opus_gnome
100A parallel corpus of GNOME localization files. Source: https://l10n.gnome.org 187 languages, 12,822 bitexts total number of files: 113,344 total number of tokens: 267.27M tot
qed_amara
78The QCRI Educational Domain Corpus (formerly QCRI AMARA Corpus) is an open multilingual collection of subtitles for educational videos and lectures collaboratively transcribed