Beam Search Strategies for Neural Machine Translation

Abstract

The basic concept in Neural Machine Translation (NMT) is to train a large Neural Network that maximizes the translation performance on a given parallel corpus. NMT is then using a simple left-to-right beam-search decoder to generate new translations that approximately maximize the trained conditional probability. The current beam search strategy generates the target sentence word by word from left-to- right while keeping a fixed amount of active candidates at each time step. First, this simple search is less adaptive as it also expands candidates whose scores are much worse than the current best. Secondly, it does not expand hypotheses if they are not within the best scoring candidates, even if their scores are close to the best one. The latter one can be avoided by increasing the beam size until no performance improvement can be observed. While you can reach better performance, this has the draw- back of a slower decoding speed. In this paper, we concentrate on speeding up the decoder by applying a more flexible beam search strategy whose candidate size may vary at each time step depending on the candidate scores. We speed up the original decoder by up to 43% for the two language pairs German-English and Chinese-English without losing any translation quality.

First Workshop on Neural Machine Translation, 2017

Cited by in corpus (15)

• Joint Admission Control and Resource Allocation of Virtual Network Embedding via Hierarchical Deep Reinforcement Learning
• Hallucination Detection in Foundation Models for Decision-Making: A Flexible Definition and Review of the State of the Art
• Beyond the C: Retargetable Decompilation using Neural Machine Translation
• Vision-based Multi-future Trajectory Prediction: A Survey
• Deanthropomorphising NLP: Can a Language Model Be Conscious?
• Towards Accurate Lip-to-Speech Synthesis in-the-Wild
• A Framework for Conditional Statement Technical Debt Identification and Description
• SuperChat: Dialogue Generation by Transfer Learning from Vision to Language using Two-dimensional Word Embedding and Pretrained ImageNet CNN Models
• Towards Validating Long-Term User Feedbacks in Interactive Recommendation Systems
• Generating Synthetic Data for Neural Keyword-to-Question Models
• Science Checker: Extractive-Boolean Question Answering For Scientific Fact Checking
• Memorization of Named Entities in Fine-tuned BERT Models
• Learning Project-wise Subsequent Code Edits via Interleaving Neural-based Induction and Tool-based Deduction
• Adaptive Beam Search to Enhance On-device Abstractive Summarization
• NeocorRAG: Less Irrelevant Information, More Explicit Evidence, and More Effective Recall via Evidence Chains