[restart] [share link] [json] select language: [swap] [compare scores] [compare models] [map] [release history] [uploads]

OPUS-MT Dashboard

Model Scores (comparing between OPUS-MT and external models)

IDBenchmark (comet)OutputOPUS-MTcometexternalcometDiff
0flores101-devtestcomparefra-eng/opus...2022-03-090.811facebook/nllb-200-3.3B0.79730.014
1flores200-devtestcompareroa-deu+eng+...2024-05-300.8902facebook/nllb-200-3.3B0.79730.093
2multi30k_test_2016_flickrcompareroa-eng/opus...2024-08-170.8868facebook/nll...illed-1.3B0.81450.072
3multi30k_test_2017_flickrcompareroa-eng/opus...2024-08-170.8926facebook/nllb-200-1.3B0.79180.101
4multi30k_test_2017_mscococompareroa-eng/opus...2024-08-170.878facebook/nll...illed-1.3B0.80570.072
5multi30k_test_2018_flickrcompareroa-eng/opus...2024-08-170.8701facebook/nll...illed-1.3B0.71850.152
6newsdiscusstest2015compareroa-eng/opus...2024-08-170.8413facebook/nll...illed-1.3B0.57280.269
7newssyscomb2009compareitc-deu+eng+...2024-05-300.8349facebook/nllb-200-3.3B0.55760.277
8newstest2009compareroa-deu+eng+...2024-05-300.8235facebook/nll...illed-1.3B0.5560.267
9newstest2010comparedeu+eng+fra+...2024-05-300.8321facebook/nllb-200-1.3B0.5940.238
10newstest2011compareroa-deu+eng+...2024-05-300.8349facebook/nllb-200-3.3B0.58330.252
11newstest2012compareroa-deu+eng+...2024-05-300.8337facebook/nllb-200-1.3B0.56490.269
12newstest2013compareitc-deu+eng+...2024-05-300.8497facebook/nllb-200-3.3B0.61590.234
13newstest2014compareroa-deu+eng+...2024-05-300.8649facebook/nll...illed-1.3B0.68780.177
14tatoeba-test-v2020-07-28compareroa-eng/opus...2024-08-170.9255facebook/nllb-200-3.3B0.9110.014
15tatoeba-test-v2021-03-30compareroa-eng/opus...2024-08-170.9255facebook/nllb-200-3.3B0.91370.012
16tatoeba-test-v2021-08-07compareroa-eng/opus...2024-08-170.9237facebook/nllb-200-3.3B0.89750.026
17tico19-testcompareroa-eng/opus...2024-08-170.8279facebook/nll...illed-1.3B0.5060.322
average0.8640.7050.159