- dataset: opus
- model: transformer
- source language(s): dan
- target language(s): swe
- model: transformer
- 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
| testset | BLEU | chr-F | #sent | #words | BP |
|---|---|---|---|---|---|
| Tatoeba-test.dan-swe | 72.1 | 0.832 | 1550 | 10082 | 0.988 |