SMOL: Professionally translated parallel data for 115 under-represented languages
This provides crucial training data for low-resource language translation, addressing a gap for many languages with no previous public resources.
The authors tackled the lack of machine translation resources for under-represented languages by releasing SMOL, a professionally translated parallel dataset for 115 languages, totaling 6.1M tokens, which improves chrF scores when used to prompt or fine-tune Large Language Models.
We open-source SMOL (Set of Maximal Overall Leverage), a suite of training data to unlock machine translation for low-resource languages. SMOL has been translated into 124 (and growing) under-resourced languages (125 language pairs), including many for which there exist no previous public resources, for a total of 6.1M translated tokens. SMOL comprises two sub-datasets, each carefully chosen for maximum impact given its size: SMOLSENT, a set of sentences chosen for broad unique token coverage, and SMOLDOC, a document-level resource focusing on a broad topic coverage. They join the already released GATITOS for a trifecta of paragraph, sentence, and token-level content. We demonstrate that using SMOL to prompt or fine-tune Large Language Models yields robust chrF improvements. In addition to translation, we provide factuality ratings and rationales for all documents in SMOLDOC, yielding the first factuality datasets for most of these languages.