MAFAND-MT: African News Machine Translation Benchmark
Largest MT benchmark for African languages in the news domain, with train/dev/test English-or-French to African-language translation pairs.
Search titles, authors, subjects and dates from across the DARE African Knowledge Repository.
25 resources found in the repository.
Largest MT benchmark for African languages in the news domain, with train/dev/test English-or-French to African-language translation pairs.
Starter notebook, benchmark guidance and contribution instructions that teach participants to build their first African-language machine-translation baseline.
Human-annotated multilingual instruction-tuning dataset curated via an open-science annotation programme with substantial African-language participation.
Large multilingual instruction dataset collection assembled for the Aya project, aggregating open datasets across 101 languages.
Massively multilingual 3B-parameter translation model covering 400 languages with broad African-language inclusion.
Crowdsourced multilingual voice corpus with clips and transcripts; large African-language subsets including FAIR Forward-supported East African languages.
Distilled 600M-parameter multilingual translation model covering 200 languages, including a large number of African languages.
Open language model adapted to African languages via continued pretraining, released by McGill-NLP.
Text-embedding model adapted for African languages and evaluated on the AfriMTEB benchmark of 59 languages and 14 tasks.
Open language model adapted to African languages via continued pretraining, released under CC BY 4.0 by McGill-NLP.
XLM-R variant with expanded vocabulary and continued pretraining for African languages, produced in African NLP research.
Open language model adapted to African languages through continued pretraining, released by McGill-NLP.