Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

MQM

Benchmarks

Task NameDataset NameSOTA ResultTrend
Span-level machine translation error detectionMQM EN-ZH annotations 2024 (test)
Precision35.95
6
Machine TranslationMQM Human Evaluation Czech→German
MQM Score10.2
3
Machine TranslationMQM Human Evaluation English→Marathi
MQM Score3.1
3
Detection of omissionsMQM gold dataset ZH-EN
Precision49.6
2
Detection of additionsMQM gold dataset ZH-EN
Precision4.3
2
Detection of omissionsMQM gold dataset EN-DE
Precision40.3
2
Showing 6 of 6 rows