[restart] [share link] 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-devtestcompareeng-fra/opus..2022-03-090.8177facebook/nllb-200-3.3B0.76510.053
1flores200-devtestcomparedeu+eng+fra+..2024-05-300.8736facebook/nllb-200-1.3B0.75810.116
2multi30k_test_2016_flickrcomparedeu+eng+fra+..2024-05-300.8577facebook/nll..illed-1.3B0.81880.039
3multi30k_test_2017_flickrcomparedeu+eng+fra+..2024-05-300.8622facebook/nll..illed-1.3B0.8070.055
4multi30k_test_2017_mscococomparedeu+eng+fra+..2024-05-300.8467facebook/nll..illed-1.3B0.79240.054
5multi30k_test_2018_flickrcomparedeu+eng+fra+..2024-05-300.8297facebook/nllb-200-3.3B0.71120.118
6newsdiscusstest2015comparedeu+eng+fra+..2024-05-300.8258facebook/nll..illed-1.3B0.60110.225
7newssyscomb2009comparedeu+eng+fra+..2024-05-300.8368facebook/nllb-200-3.3B0.59850.238
8newstest2008showeng-fra/opus..2022-03-090.82310.823
9newstest2009comparedeu+eng+fra+..2024-05-300.8244facebook/nllb-200-1.3B0.57610.248
10newstest2010comparedeu+eng+fra+..2024-05-300.8332facebook/nll..illed-1.3B0.62760.206
11newstest2011comparedeu+eng+fra+..2024-05-300.8305facebook/nllb-200-3.3B0.62130.209
12newstest2012comparedeu+eng+fra+..2024-05-300.8291facebook/nll..illed-1.3B0.58490.244
13newstest2013comparedeu+eng+fra+..2024-05-300.8444facebook/nll..illed-1.3B0.63340.211
14newstest2014comparedeu+eng+fra+..2024-05-300.863facebook/nll..illed-1.3B0.73220.131
15ntrex128showeng-fra/opus..2022-03-090.83450.835
16tatoeba-test-v2020-07-28compareeng-fra/opus..2022-03-090.9035facebook/nll..illed-1.3B0.88130.022
17tatoeba-test-v2021-03-30compareeng-fra/opus..2022-03-090.9107facebook/nll..illed-1.3B0.88680.024
18tatoeba-test-v2021-08-07compareeng-fra/opus..2022-03-090.9059facebook/nll..illed-1.3B0.87910.027
19tico19-testcomparedeu+eng+fra+..2024-05-300.8049facebook/nll..illed-1.3B0.48460.320
average0.8480.7090.139