opusTCv20210807_transformer-big_2022-07-23.zip dataset: opusTCv20210807 model: transformer-big source language(s): fra ita lat_Latn mwl pob por spa target language(s): eus raw source language(s): fra ita lat mwl pob por spa raw target language(s): eus model: transformer-big pre-processing: normalization + SentencePiece (spm32k,spm32k) download: opusTCv20210807_transformer-big_2022-07-23.zip test set translations: opusTCv20210807_transformer-big_2022-07-23.test.txt test set scores: opusTCv20210807_transformer-big_2022-07-23.eval.txt Benchmarks testset BLEU chr-F #sent #words BP Tatoeba-test-v2021-08-07.fra-eus 27.2 0.53646 144 811 0.989 Tatoeba-test-v2021-08-07.ita-eus 29.5 0.61906 62 351 0.974 Tatoeba-test-v2021-08-07.lat-eus 3.4 0.22748 1 17 1.000 Tatoeba-test-v2021-08-07.multi-eus 32.1 0.60335 2058 12116 0.983 Tatoeba-test-v2021-08-07.mwl-eus 19.0 0.29796 1 4 1.000 Tatoeba-test-v2021-08-07.por-eus 100.0 10.00000 1 5 1.000 Tatoeba-test-v2021-08-07.spa-eus 32.2 0.60500 1850 10939 0.983