opus-2021-02-18.zip

Benchmarks

testset	BLEU	chr-F	#sent	#words	BP
Tatoeba-test.ukr-deu	48.2	0.661	10000	62298	0.980

dataset: opusTCv20210807+pft
model: transformer-align
source language(s): ukr
target language(s): deu
raw source language(s): ukr
raw target language(s): deu
model: transformer-align
pre-processing: normalization + SentencePiece (spm32k,spm32k)
download: opusTCv20210807+pft_transformer-align_2022-03-06.zip
test set translations: opusTCv20210807+pft_transformer-align_2022-03-06.test.txt
test set scores: opusTCv20210807+pft_transformer-align_2022-03-06.eval.txt

testset	BLEU	chr-F	#sent	#words	BP
Tatoeba-test-v2021-08-07.ukr-deu	52.2	0.69211	10000	62652	0.991

dataset: opusTCv20210807+pft
model: transformer-tiny11-align
source language(s): ukr
target language(s): deu
raw source language(s): ukr
raw target language(s): deu
model: transformer-tiny11-align
pre-processing: normalization + SentencePiece (spm32k,spm32k)
download: opusTCv20210807+pft-sepvoc_transformer-tiny11-align_2022-03-16.zip
test set translations: opusTCv20210807+pft-sepvoc_transformer-tiny11-align_2022-03-16.test.txt
test set scores: opusTCv20210807+pft-sepvoc_transformer-tiny11-align_2022-03-16.eval.txt

testset	BLEU	chr-F	#sent	#words	BP
Tatoeba-test-v2021-08-07.ukr-deu	49.0	0.66981	10000	62652	0.988

dataset: opusTCv20210807+nopar+ft95
model: transformer-tiny11-align
source language(s): ukr
target language(s): deu
raw source language(s): ukr
raw target language(s): deu
model: transformer-tiny11-align
pre-processing: normalization + SentencePiece (spm32k,spm32k)
download: opusTCv20210807+nopar+ft95-sepvoc_transformer-tiny11-align_2022-03-21.zip
test set translations: opusTCv20210807+nopar+ft95-sepvoc_transformer-tiny11-align_2022-03-21.test.txt
test set scores: opusTCv20210807+nopar+ft95-sepvoc_transformer-tiny11-align_2022-03-21.eval.txt

testset	BLEU	chr-F	#sent	#words	BP
Tatoeba-test-v2021-08-07.ukr-deu	52.0	0.69033	10000	62661	0.983