Unsupervised Machine Translation On Wmt2014 2
Métriques
BLEU
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | BLEU |
|---|---|
| language-models-are-few-shot-learners | 32.6 |
| mass-masked-sequence-to-sequence-pre-training | 37.5 |
| phrase-based-neural-unsupervised-machine | 27.6 |
| incorporating-bert-into-neural-machine-1 | 38.27 |
| cross-lingual-language-model-pretraining | 33.4 |
| an-effective-approach-to-unsupervised-machine | 36.2 |
| unsupervised-neural-machine-translation-with | 29.5 |