opus-2021-02-19.zip

dataset: opus
model: transformer-align
source language(s): ukr
target language(s): nld
model: transformer-align
pre-processing: normalization + SentencePiece (spm32k,spm32k)
download: opus-2021-02-19.zip
test set translations: opus-2021-02-19.test.txt
test set scores: opus-2021-02-19.eval.txt

Benchmarks

testset	BLEU	chr-F	#sent	#words	BP
Tatoeba-test.ukr-nld	48.7	0.656	10000	59943	0.985