论文标题
使用单语言数据改善非自动入学的神经机器翻译
Improving Non-autoregressive Neural Machine Translation with Monolingual Data
论文作者
论文摘要
非自动回旋（NAR）神经机译通常是通过自回归（AR）模型的知识蒸馏而进行的。在此框架下，我们利用大型单语言语料库来提高NAR模型的性能，目的是在防止过度拟合的同时传递AR模型的概括能力。除了强大的NAR基线外，我们对WMT14 EN-DE和WMT16 ENRO新闻翻译任务的实验结果确认，单语言数据的增强始终提高NAR模型的性能，以取得与教师AR模型的性能相比或更好的结果，而不是与最佳的NAR NAR方法相比或更好的结果，这有助于减少文献中的NAR方法，从而有助于培训过程。
Non-autoregressive (NAR) neural machine translation is usually done via knowledge distillation from an autoregressive (AR) model. Under this framework, we leverage large monolingual corpora to improve the NAR model's performance, with the goal of transferring the AR model's generalization ability while preventing overfitting. On top of a strong NAR baseline, our experimental results on the WMT14 En-De and WMT16 En-Ro news translation tasks confirm that monolingual data augmentation consistently improves the performance of the NAR model to approach the teacher AR model's performance, yields comparable or better results than the best non-iterative NAR methods in the literature and helps reduce overfitting in the training process.
