opus-2020-10-04.zip dataset: opus model: transformer source language(s): eng jpn target language(s): eng jpn 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) download: opus-2020-10-04.zip test set translations: opus-2020-10-04.test.txt test set scores: opus-2020-10-04.eval.txt Benchmarks testset BLEU chr-F Tatoeba-test.eng-jpn.eng.jpn 11.3 0.230 Tatoeba-test.jpn-eng.jpn.eng 25.5 0.461 Tatoeba-test.multi.multi 19.6 0.386