Understanding and Improving Lexical Choice in Non-Autoregressive Translation
Mai 3, 2021
Sprecher:innen
Liang Ding
Sprecher:in · 1 Follower:in
Longyue Wang
Sprecher:in · 1 Follower:in
Derek F. Wong
Sprecher:in · 0 Follower:innen
Dacheng Tao
Sprecher:in · 1 Follower:in
Zhaopeng Tu
Sprecher:in · 2 Follower:innen
Über
Knowledge distillation (KD) is essential for training non-autoregressive translation (NAT) models by reducing the complexity of the raw data with an autoregressive teacher model. In this study, we empirically show that as a side effect of this training, the lexical choice errors on low-frequency words are propagated to the NAT model from the teacher model. To alleviate this problem, we propose to expose the raw data to NAT models to restore the useful information of low-frequency words, which are missed in the distilled data. To this end, we introduce an extra Kullback-Leibler divergence term derived by comparing the lexical choice of NAT model and that embedded in the raw data. Experimental results across language pairs and model architectures demonstrate the effectiveness and universality of the proposed approach. Extensive analyses confirm our claim that our approach improves performance by reducing the lexical choice errors on low-frequency words. Encouragingly, our approach pushes the SOTA NAT performance on the WMT14 English-German and WMT16 Romanian-English datasets up to 27.8 and 33.8 BLEU points, respectively. The codes and models will be released.Knowledge distillation (KD) is essential for training non-autoregressive translation (NAT) models by reducing the complexity of the raw data with an autoregressive teacher model. In this study, we empirically show that as a side effect of this training, the lexical choice errors on low-frequency words are propagated to the NAT model from the teacher model. To alleviate this problem, we propose to expose the raw data to NAT models to restore the useful information of low-frequency words, which ar…
Über ICLR 2021
The International Conference on Learning Representations (ICLR) is the premier gathering of professionals dedicated to the advancement of the branch of artificial intelligence called representation learning, but generally referred to as deep learning. ICLR is globally renowned for presenting and publishing cutting-edge research on all aspects of deep learning used in the fields of artificial intelligence, statistics and data science, as well as important application areas such as machine vision, computational biology, speech recognition, text understanding, gaming, and robotics.
Gefällt euch das Format? Vertraut auf SlidesLive, um euer nächstes Event festzuhalten!
Professionelle Aufzeichnung und Livestreaming – weltweit.
Freigeben
Empfohlene Videos
Präsentationen, deren Thema, Kategorie oder Sprecher:in ähnlich sind
SIMPLE SPECTRAL GRAPH CONVOLUTION
Später ansehen
Implicit Normalizing Flows
Später ansehen
Cheng Lu, …
Learning Invariant Representations for Reinforcement Learning without Reconstruction
Später ansehen
Amy Zhang, …
Adversarial attacks on models for computer programs
Später ansehen
A Machine Learning Model for Predicting Deterioration of COVID-19 Inpatients
Später ansehen
Omer Noy, …
