bat-bat

opus-2021-02-18.zip

• dataset: opus
• model: transformer
• source language(s): lav lit ltg
• target language(s): lav lit ltg
• model: transformer
• 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: >>lav<< >>eng<< >>lit<< >>ltg<<
• download: opus-2021-02-18.zip
• test set translations: opus-2021-02-18.test.txt
• test set scores: opus-2021-02-18.eval.txt

Benchmarks

testset	BLEU	chr-F	#sent	#words	BP
newsdev2017-enlv.eng-lav	21.6	0.521	2003	41503	0.992
newsdev2017-enlv.lav-eng	24.8	0.549	2003	48175	1.000
newsdev2019-enlt.eng-lit	18.5	0.509	2000	40116	1.000
newsdev2019-enlt.lit-eng	25.0	0.535	2000	49666	0.974
newstest2017-enlv.eng-lav	16.0	0.480	2001	39434	1.000
newstest2017-enlv.lav-eng	19.2	0.494	2001	47511	1.000
newstest2019-enlt.eng-lit	12.5	0.451	998	20091	1.000
newstest2019-lten.lit-eng	27.8	0.574	1000	26079	0.958
Tatoeba-test.lav-lav	11.2	0.349	8	38	1.000
Tatoeba-test.lav-lit	24.3	0.712	2	11	1.000
Tatoeba-test.lav-ltg	4.8	0.241	4	19	1.000
Tatoeba-test.lit-lav	26.1	0.780	2	11	1.000
Tatoeba-test.ltg-lav	24.6	0.468	4	19	1.000
Tatoeba-test.multi-multi	41.6	0.611	8812	57287	0.978