OPUS-MT-train/models/en+fr-et+hu+fi

opus-2019-12-04.zip

• dataset: opus
• model: transformer
• pre-processing: normalization + tokenization + BPE
• a sentence initial language token is required in the form of >>id<< (id = valid target language ID)
• download: opus-2019-12-04.zip
• test set translations: opus-2019-12-04.test.txt
• test set scores: opus-2019-12-04.eval.txt

Benchmarks

testset	BLEU	chr-F
newsdev2015-enfi.en.fi	17.6	0.514
newsdev2018-enet.en.et	18.8	0.504
newstest2015-enfi.en.fi	19.1	0.522
newstest2016-enfi.en.fi	20.1	0.536
newstest2017-enfi.en.fi	22.7	0.558
newstest2018-enet.en.et	19.4	0.512
newstest2018-enfi.en.fi	15.5	0.489
newstest2019-enfi.en.fi	19.4	0.507
newstestB2016-enfi.en.fi	16.3	0.504
newstestB2017-enfi.en.fi	18.1	0.520
Tatoeba.en.fi	36.7	0.604