ukr-eng

opus-2021-02-19.zip

• dataset: opus
• model: transformer-align
• source language(s): ukr
• target language(s): eng
• 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-eng	51.7	0.670	10000	66118	0.972

opus+bt-2021-04-30.zip

• dataset: opus+bt
• model: transformer-align
• source language(s): ukr
• target language(s): eng
• model: transformer-align
• pre-processing: normalization + SentencePiece (spm32k,spm32k)
• download: opus+bt-2021-04-30.zip
• test set translations: opus+bt-2021-04-30.test.txt
• test set scores: opus+bt-2021-04-30.eval.txt

Benchmarks

testset	BLEU	chr-F	#sent	#words	BP
Tatoeba-test.ukr-eng	52.2	0.675	10000	66118	0.977

opusTCv20210807+nopar+ft95-sepvoc_transformer-tiny11-align_2022-03-05.zip

• dataset: opusTCv20210807+nopar+ft95
• model: transformer-tiny11-align
• source language(s): ukr
• target language(s): eng
• raw source language(s): ukr
• raw target language(s): eng
• model: transformer-tiny11-align
• pre-processing: normalization + SentencePiece (spm32k,spm32k)
• download: opusTCv20210807+nopar+ft95-sepvoc_transformer-tiny11-align_2022-03-05.zip
• test set translations: opusTCv20210807+nopar+ft95-sepvoc_transformer-tiny11-align_2022-03-05.test.txt
• test set scores: opusTCv20210807+nopar+ft95-sepvoc_transformer-tiny11-align_2022-03-05.eval.txt

Benchmarks

testset	BLEU	chr-F	#sent	#words	BP
Tatoeba-test-v2021-08-07.intgemm8.shortlist.ukr-eng	50.1	0.66250	10000	67543	0.981
Tatoeba-test-v2021-08-07.intgemm8tunedshiftAlphas.shortlist.ukr-eng	49.6	0.65743	10000	67543	0.979
Tatoeba-test-v2021-08-07.intgemm8tunedshiftAlphas.ukr-eng	49.6	0.65763	10000	67543	0.979
Tatoeba-test-v2021-08-07.intgemm8.ukr-eng	50.2	0.66282	10000	67543	0.981
Tatoeba-test-v2021-08-07.ukr-eng	51.4	0.67210	10000	67543	0.981

opusTCv20210807+nopar+ft95-sepvoc_transformer-tiny11-align_2022-03-23.zip

• dataset: opusTCv20210807+nopar+ft95
• model: transformer-tiny11-align
• source language(s): ukr
• target language(s): eng
• raw source language(s): ukr
• raw target language(s): eng
• model: transformer-tiny11-align
• pre-processing: normalization + SentencePiece (spm32k,spm32k)
• download: opusTCv20210807+nopar+ft95-sepvoc_transformer-tiny11-align_2022-03-23.zip
• test set translations: opusTCv20210807+nopar+ft95-sepvoc_transformer-tiny11-align_2022-03-23.test.txt
• test set scores: opusTCv20210807+nopar+ft95-sepvoc_transformer-tiny11-align_2022-03-23.eval.txt

Benchmarks

testset	BLEU	chr-F	#sent	#words	BP
flores101.ukr-eng	35.8	0.62527	1012	24739	1.000
Tatoeba-test-v2021-08-07.ukr-eng	52.7	0.68333	10000	67543	0.977