DeepAI AI Chat
Log In Sign Up

Quality-Aware Decoding for Neural Machine Translation

by   Patrick Fernandes, et al.
University of Lisbon
Carnegie Mellon University

Despite the progress in machine translation quality estimation and evaluation in the last years, decoding in neural machine translation (NMT) is mostly oblivious to this and centers around finding the most probable translation according to the model (MAP decoding), approximated with beam search. In this paper, we bring together these two lines of research and propose quality-aware decoding for NMT, by leveraging recent breakthroughs in reference-free and reference-based MT evaluation through various inference methods like N-best reranking and minimum Bayes risk decoding. We perform an extensive comparison of various possible candidate generation and ranking methods across four datasets and two model classes and find that quality-aware decoding consistently outperforms MAP-based decoding according both to state-of-the-art automatic metrics (COMET and BLEURT) and to human assessments. Our code is available at


page 16

page 17


Neural Machine Translation: Challenges, Progress and Future

Machine translation (MT) is a technique that leverages computers to tran...

RMBR: A Regularized Minimum Bayes Risk Reranking Framework for Machine Translation

Beam search is the most widely used decoding method for neural machine t...

UniTE: Unified Translation Evaluation

Translation quality evaluation plays a crucial role in machine translati...

Is MAP Decoding All You Need? The Inadequacy of the Mode in Neural Machine Translation

Recent studies have revealed a number of pathologies of neural machine t...

Adam Mickiewicz University at WMT 2022: NER-Assisted and Quality-Aware Neural Machine Translation

This paper presents Adam Mickiewicz University's (AMU) submissions to th...

Competency-Aware Neural Machine Translation: Can Machine Translation Know its Own Translation Quality?

Neural machine translation (NMT) is often criticized for failures that h...

Minimum Bayes Risk Decoding with Neural Metrics of Translation Quality

This work applies Minimum Bayes Risk (MBR) decoding to optimize diverse ...