- dataset: opus+bt
- model: transformer-align
- source language(s): eng
- target language(s): fry
- model: transformer-align
- pre-processing: normalization + SentencePiece (spm4k,spm4k)
- download: opus+bt-2021-03-12.zip
- test set translations: opus+bt-2021-03-12.test.txt
- test set scores: opus+bt-2021-03-12.eval.txt
| testset | BLEU | chr-F | #sent | #words | BP |
|---|---|---|---|---|---|
| Tatoeba-test.eng-fry | 6.1 | 0.220 | 205 | 1529 | 0.834 |