- dataset: opus1m
- model: transformer-align
- source language(s): bel orv
- target language(s): rus
- model: transformer-align
- pre-processing: normalization + SentencePiece (spm32k,spm32k)
- download: opus1m-2021-05-16.zip
- test set translations: opus1m-2021-05-16.test.txt
- test set scores: opus1m-2021-05-16.eval.txt
| testset | BLEU | chr-F | #sent | #words | BP |
|---|---|---|---|---|---|
| Tatoeba-test.bel_Latn-rus | 1.2 | 0.010 | 6 | 60 | 1.000 |
| Tatoeba-test.bel-rus | 44.1 | 0.638 | 2500 | 18815 | 1.000 |
| Tatoeba-test.multi-rus | 41.8 | 0.614 | 2671 | 20161 | 0.999 |
| Tatoeba-test.orv-rus | 0.5 | 0.153 | 171 | 1259 | 0.965 |
| Tatoeba-test.ukr-rus | 4.5 | 0.209 | 10000 | 60129 | 0.948 |