opus-2021-02-23.zip dataset: opus model: transformer source language(s): heb target language(s): apc ara arq arz model: transformer pre-processing: normalization + SentencePiece (spm32k,spm32k) a sentence initial language token is required in the form of >>id<< (id = valid target language ID) valid language labels: >>ara<< >>arq<< >>arz<< download: opus-2021-02-23.zip test set translations: opus-2021-02-23.test.txt test set scores: opus-2021-02-23.eval.txt Benchmarks testset BLEU chr-F #sent #words BP Tatoeba-test.heb-apc 10.7 0.185 1 4 1.000 Tatoeba-test.heb-ara 23.6 0.532 1208 6372 0.926 Tatoeba-test.heb-arq 0.0 0.514 1 4 0.717 Tatoeba-test.heb-arz 7.6 0.305 2 8 1.000