Alleviating repetitive tokens in non-autoregressive machine translation with unlikelihood training.
Shuheng WangShumin ShiHeyan HuangPublished in: Soft Comput. (2024)
Keyphrases
- machine translation
- autoregressive
- non stationary
- moving average
- language independent
- gaussian markov random field
- natural language processing
- cross lingual
- information extraction
- chinese english
- cross language information retrieval
- word alignment
- machine translation system
- random fields
- natural language
- target language
- language resources
- statistical machine translation
- parallel corpus
- information retrieval
- brazilian portuguese
- finite state transducers
- multiresolution
- multiscale