Breaking the Beam Search Curse: A Study of (Re-)Scoring Methods and Stopping Criteria for Neural Machine Translation
arXiv:1808.09582
Abstract
Beam search is widely used in neural machine translation, and usually improves translation quality compared to greedy search. It has been widely observed that, however, beam sizes larger than 5 hurt translation quality. We explain why this happens, and propose several methods to address this problem. Furthermore, we discuss the optimal stopping criteria for these methods. Results show that our hyperparameter-free methods outperform the widely-used hyperparameter-free heuristic of length normalization by +2.0 BLEU, and achieve the best results among all methods on Chinese-to-English translation.
arXiv commentEMNLP 2018
References in corpus (6)
- Sequence to Sequence Learning with Neural Networks
- Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
- Convolutional Sequence to Sequence Learning
- Sequence-to-Sequence Learning as Beam-Search Optimization
- Neural Machine Translation with Reconstruction
- Analyzing Uncertainty in Neural Machine Translation