| ID | Language | Benchmark | Output | bleu |
|---|---|---|---|---|
| 0 | eng-spa | flores101-devtest | show | 27.2 |
| 1 | eng-spa | flores200-devtest | show | 27.3 |
| 2 | eng-spa | newssyscomb2009 | show | 31.8 |
| 3 | eng-spa | newstest2008 | show (logfile) | 29.9 |
| 4 | eng-spa | newstest2009 | show | 30.8 |
| 5 | eng-spa | newstest2010 | show | 37.2 |
| 6 | eng-spa | newstest2011 | show | 39.2 |
| 7 | eng-spa | newstest2012 | show | 39.4 |
| 8 | eng-spa | newstest2013 | show | 36 |
| 9 | eng-spa | ntrex128 | show (logfile) | 41.4 |
| 10 | eng-spa | tatoeba-test-v2020-07-28 | show | 54.3 |
| 11 | eng-spa | tatoeba-test-v2021-03-30 | show | 55.1 |
| 12 | eng-spa | tatoeba-test-v2021-08-07 | show | 56.1 |
| 13 | eng-spa | tico19-test | show | 54.1 |
| show all | average | 39.986 |