- dataset: opusTCv20210807+bt
- model: transformer-big
- source language(s): cat oci spa
- target language(s): eng
- raw source language(s): cat oci spa
- raw target language(s): eng
- model: transformer-big
- pre-processing: normalization + SentencePiece (spm32k,spm32k)
- download: opusTCv20210807+bt_transformer-big_2022-03-13.zip
- test set translations: opusTCv20210807+bt_transformer-big_2022-03-13.test.txt
- test set scores: opusTCv20210807+bt_transformer-big_2022-03-13.eval.txt
| testset | BLEU | chr-F | #sent | #words | BP |
|---|---|---|---|---|---|
| newssyscomb2009.spa-eng | 30.8 | 0.57104 | 502 | 11821 | 0.987 |
| news-test2008.spa-eng | 27.9 | 0.55431 | 2051 | 49380 | 0.990 |
| newstest2009.spa-eng | 30.2 | 0.57153 | 2525 | 65402 | 0.975 |
| newstest2010.spa-eng | 36.8 | 0.61883 | 2489 | 61724 | 0.985 |
| newstest2011.spa-eng | 34.7 | 0.60278 | 3003 | 74681 | 0.990 |
| newstest2012.spa-eng | 38.6 | 0.62773 | 3003 | 72812 | 1.000 |
| newstest2013.spa-eng | 35.3 | 0.60994 | 3000 | 64505 | 1.000 |
| Tatoeba-test-v2021-08-07.cat-eng | 56.9 | 0.71799 | 1631 | 12625 | 0.982 |
| Tatoeba-test-v2021-08-07.multi-eng | 60.7 | 0.74626 | 10000 | 82335 | 0.983 |
| Tatoeba-test-v2021-08-07.oci-eng | 19.3 | 0.36347 | 841 | 5299 | 1.000 |
| Tatoeba-test-v2021-08-07.spa-eng | 62.1 | 0.75914 | 16583 | 138085 | 0.979 |
| tico19-test.spa-eng | 51.8 | 0.74048 | 2100 | 56339 | 0.992 |