WMT
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
WMT all 54 LPs 24++
59.85SLI
11
WMT EN-ET (test)
12.6BLEU
11
WMT EN-FI (test)
9.4BLEU
11
WMT EN-TR (test)
11.7BLEU
11
WMT ET-EN (test)
16.8BLEU
11
WMT FI-EN (test)
15BLEU
11
WMT TR-EN (test)
18.5BLEU
11
WMT RU-EN (test)
22.2BLEU
11
WMT CS-EN (test)
21.6BLEU
11
WMT en→is (test)
26.68BLEU
11
WMT en→cs (test)
0.376BLEU
11
WMT17 Zh-En (test)
26.5BLEU
11
WMT En-Ro 16
37.7BLEU
11
WMT En-Lv 17
24.4BLEU
11
WMT En-Fi 17
28.7BLEU
11
WMT En-Et 18
28.5BLEU
11
WMT En-Kk 19
12.9BLEU
11
WMT English-German (EN-DE) 2014 (test)
30.1BLEU Score
11
WMT En-De 16 full (test)
30.1BLEU
11
WMT'15 German-English (test)
32.1BLEU
11
WMT Metrics Shared Task EN–DE 2023
0.83Average Correlation
10
WMT En->Zh
67.21BLEURT-20
10
WMT En-Fr Law OOD ID 15 (test)
65.01AUROC
10
WMT En-De 16
49.86chrF
10
WMT En-Fr 14
55.82chrF
10