FLoRes Benchmark

The FLoRes dataset is a benchmark designed for low-resource machine translation. It includes English-to-Nepali translations with approximately 564,000 parallel sentences, making it considerably more challenging due to the language distance.