| ID | Language | Benchmark | Output | bleu |
|---|---|---|---|---|
| 0 | eng-ces | flores101-devtest | show | 29.2 |
| 1 | eng-ces | multi30k_test_2016_flickr | show | 32.4 |
| 2 | eng-ces | multi30k_test_2018_flickr | show | 32.2 |
| 3 | eng-ces | newssyscomb2009 | show | 22.9 |
| 4 | eng-ces | newstest2009 | show | 21.6 |
| 5 | eng-ces | newstest2010 | show | 21.5 |
| 6 | eng-ces | newstest2011 | show | 23.1 |
| 7 | eng-ces | newstest2012 | show | 20.5 |
| 8 | eng-ces | newstest2013 | show | 24.3 |
| 9 | eng-ces | newstest2014 | show | 27.3 |
| 10 | eng-ces | newstest2015 | show | 23.7 |
| 11 | eng-ces | newstest2016 | show | 26.9 |
| 12 | eng-ces | newstest2017 | show | 21.6 |
| 13 | eng-ces | newstest2018 | show | 21.1 |
| 14 | eng-ces | newstest2019 | show | 22.4 |
| 15 | eng-ces | tatoeba-test-v2020-07-28 | show | 48.2 |
| 16 | eng-ces | tatoeba-test-v2021-03-30 | show | 48 |
| 17 | eng-ces | tatoeba-test-v2021-08-07 | show | 47.8 |
| average | 28.594 |