| ID | Language | Benchmark | Output | bleu |
|---|---|---|---|---|
| 0 | eng-fin | flores101-devtest | show (logfile) | 23 |
| 1 | eng-fin | flores200-devtest | show (logfile) | 23 |
| 2 | eng-fin | newstest2015 | show (logfile) | 22.5 |
| 3 | eng-fin | newstest2016 | show (logfile) | 24.6 |
| 4 | eng-fin | newstest2017 | show (logfile) | 27 |
| 5 | eng-fin | newstest2018 | show (logfile) | 16.7 |
| 6 | eng-fin | newstest2019 | show (logfile) | 23.1 |
| 7 | eng-fin | newstestALL2016 | show (logfile) | 27.9 |
| 8 | eng-fin | newstestALL2017 | show (logfile) | 30.7 |
| 9 | eng-fin | newstestB2016 | show (logfile) | 19.6 |
| 10 | eng-fin | newstestB2017 | show (logfile) | 22.4 |
| 11 | eng-fin | ntrex128 | show (logfile) | 17.6 |
| 12 | eng-fin | tatoeba-test-v2020-07-28 | show (logfile) | 38.8 |
| 13 | eng-fin | tatoeba-test-v2021-03-30 | show (logfile) | 38.7 |
| 14 | eng-fin | tatoeba-test-v2021-08-07 | show (logfile) | 38.6 |
| average | 26.280 |