deu-zle

opusTCv20210807_transformer-big_2022-03-13.zip

• dataset: opusTCv20210807
• model: transformer-big
• source language(s): deu
• target language(s): bel bel_Latn orv_Cyrl rus ukr
• raw source language(s): deu
• raw target language(s): bel orv rus ukr
• model: transformer-big
• pre-processing: normalization + SentencePiece (spm32k,spm32k)
• a sentence initial language token is required in the form of >>id<< (id = valid target language ID)
• valid language labels:
• download: opusTCv20210807_transformer-big_2022-03-13.zip
• test set translations: opusTCv20210807_transformer-big_2022-03-13.test.txt
• test set scores: opusTCv20210807_transformer-big_2022-03-13.eval.txt

Benchmarks

testset	BLEU	chr-F	#sent	#words	BP
newstest2012.deu-rus	20.0	0.48631	3003	64830	0.991
newstest2013.deu-rus	23.9	0.51726	3000	58560	0.969
Tatoeba-test-v2021-08-07.deu-bel	25.2	0.49854	551	3596	1.000
Tatoeba-test-v2021-08-07.deu-bel_Latn	3.1	0.696	3	21	1.000
Tatoeba-test-v2021-08-07.deu-multi	43.3	0.64853	10000	61925	0.993
Tatoeba-test-v2021-08-07.deu-orv	0.8	0.16018	28	139	1.000
Tatoeba-test-v2021-08-07.deu-rus	43.6	0.65387	12800	86919	0.992
Tatoeba-test-v2021-08-07.deu-ukr	39.0	0.61380	10319	56121	1.000

opusTCv20210807_transformer-big_2022-03-19.zip

• dataset: opusTCv20210807
• model: transformer-big
• source language(s): deu
• target language(s): bel bel_Latn orv_Cyrl rus ukr
• raw source language(s): deu
• raw target language(s): bel orv rus ukr
• model: transformer-big
• pre-processing: normalization + SentencePiece (spm32k,spm32k)
• a sentence initial language token is required in the form of >>id<< (id = valid target language ID)
• valid language labels:
• download: opusTCv20210807_transformer-big_2022-03-19.zip
• test set translations: opusTCv20210807_transformer-big_2022-03-19.test.txt
• test set scores: opusTCv20210807_transformer-big_2022-03-19.eval.txt

Benchmarks

testset	BLEU	chr-F	#sent	#words	BP
newstest2012.deu-rus	20.6	0.49401	3003	64830	0.983
newstest2013.deu-rus	24.7	0.52514	3000	58560	0.963
Tatoeba-test-v2021-08-07.deu-bel	28.5	0.52805	551	3596	1.000
Tatoeba-test-v2021-08-07.deu-bel_Latn	3.1	0.694	3	21	1.000
Tatoeba-test-v2021-08-07.deu-multi	43.9	0.65226	10000	61925	0.988
Tatoeba-test-v2021-08-07.deu-orv	0.7	0.16532	28	139	1.000
Tatoeba-test-v2021-08-07.deu-rus	45.2	0.66607	12800	86919	0.986
Tatoeba-test-v2021-08-07.deu-ukr	40.6	0.62749	10319	56121	0.999

opusTCv20210807_transformer-big_2022-03-23.zip

• dataset: opusTCv20210807
• model: transformer-big
• source language(s): deu
• target language(s): bel bel_Latn orv_Cyrl rus ukr
• raw source language(s): deu
• raw target language(s): bel orv rus ukr
• model: transformer-big
• pre-processing: normalization + SentencePiece (spm32k,spm32k)
• a sentence initial language token is required in the form of >>id<< (id = valid target language ID)
• valid language labels:
• download: opusTCv20210807_transformer-big_2022-03-23.zip
• test set translations: opusTCv20210807_transformer-big_2022-03-23.test.txt
• test set scores: opusTCv20210807_transformer-big_2022-03-23.eval.txt

Benchmarks

testset	BLEU	chr-F	#sent	#words	BP
newstest2012.deu-rus	20.8	0.49414	3003	64830	0.979
newstest2013.deu-rus	24.9	0.52632	3000	58560	0.961
Tatoeba-test-v2021-08-07.deu-bel	29.2	0.53063	551	3596	1.000
Tatoeba-test-v2021-08-07.deu-bel_Latn	3.1	0.694	3	21	1.000
Tatoeba-test-v2021-08-07.deu-multi	43.8	0.65199	10000	61925	0.987
Tatoeba-test-v2021-08-07.deu-orv	0.7	0.16833	28	139	1.000
Tatoeba-test-v2021-08-07.deu-rus	45.2	0.66714	12800	86919	0.986
Tatoeba-test-v2021-08-07.deu-ukr	40.5	0.62641	10319	56121	0.998