The <scp>Flores-101</scp> Evaluation Benchmark for Low-Resource and Multilingual Machine Translation

TLDR

Low‑resource and multilingual machine translation lacks good evaluation benchmarks, which either have limited coverage, restricted domains, or low quality due to semi‑automatic construction. This work introduces the Flores‑101 benchmark of 3001 Wikipedia sentences covering diverse topics and publicly releases it to advance machine translation research. The benchmark consists of 3001 sentences translated into 101 languages by professional translators in a carefully controlled process. The dataset allows better assessment of model quality on low‑resource languages and many‑to‑many multilingual systems, with fully aligned translations.

Abstract

Abstract One of the biggest challenges hindering progress in low-resource and multilingual machine translation is the lack of good evaluation benchmarks. Current evaluation benchmarks either lack good coverage of low-resource languages, consider only restricted domains, or are low quality because they are constructed using semi-automatic procedures. In this work, we introduce the Flores-101 evaluation benchmark, consisting of 3001 sentences extracted from English Wikipedia and covering a variety of different topics and domains. These sentences have been translated in 101 languages by professional translators through a carefully controlled process. The resulting dataset enables better assessment of model quality on the long tail of low-resource languages, including the evaluation of many-to-many multilingual translation systems, as all translations are fully aligned. By publicly releasing such a high-quality and high-coverage dataset, we hope to foster progress in the machine translation community and beyond.