| ID | Language | Benchmark | Output | bleu |
|---|---|---|---|---|
| 0 | eng-fin | flores101-devtest | show | 27.4 |
| 1 | eng-fin | flores200-devtest | show | 27.4 |
| 2 | eng-fin | newstest2015 | show | 26.9 |
| 3 | eng-fin | newstest2016 | show | 28.7 |
| 4 | eng-fin | newstest2017 | show | 31.7 |
| 5 | eng-fin | newstest2018 | show | 20 |
| 6 | eng-fin | newstest2019 | show | 26.8 |
| 7 | eng-fin | newstestALL2016 | show | 32.6 |
| 8 | eng-fin | newstestALL2017 | show | 36.3 |
| 9 | eng-fin | newstestB2016 | show | 23.2 |
| 10 | eng-fin | newstestB2017 | show | 27.3 |
| 11 | eng-fin | ntrex128 | show | 20.6 |
| 12 | eng-fin | tatoeba-test-v2020-07-28 | show | 38.8 |
| 13 | eng-fin | tatoeba-test-v2021-03-30 | show | 38.8 |
| 14 | eng-fin | tatoeba-test-v2021-08-07 | show | 38.8 |
| average | 29.687 |