@inproceedings{zhu-etal-2022-non,
title = "Non-Autoregressive Neural Machine Translation with Consistency Regularization Optimized Variational Framework",
author = "Zhu, Minghao and
Wang, Junli and
Yan, Chungang",
editor = "Carpuat, Marine and
de Marneffe, Marie-Catherine and
Meza Ruiz, Ivan Vladimir",
booktitle = "Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies",
month = jul,
year = "2022",
address = "Seattle, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.naacl-main.45",
doi = "10.18653/v1/2022.naacl-main.45",
pages = "607--617",
abstract = "Variational Autoencoder (VAE) is an effective framework to model the interdependency for non-autoregressive neural machine translation (NAT). One of the prominent VAE-based NAT frameworks, LaNMT, achieves great improvements to vanilla models, but still suffers from two main issues which lower down the translation quality: (1) mismatch between training and inference circumstances and (2) inadequacy of latent representations. In this work, we target on addressing these issues by proposing posterior consistency regularization. Specifically, we first perform stochastic data augmentation on the input samples to better adapt the model for inference circumstance, and then conduct consistency training on posterior latent variables to construct a more robust latent representations without any expansion on latent size. Experiments on En{\textless}-{\textgreater}De and En{\textless}-{\textgreater}Ro benchmarks confirm the effectiveness of our methods with about 1.5/0.7 and 0.8/0.3 BLEU points improvement to the baseline model with about $12.6\times$ faster than autoregressive Transformer.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhu-etal-2022-non">
<titleInfo>
<title>Non-Autoregressive Neural Machine Translation with Consistency Regularization Optimized Variational Framework</title>
</titleInfo>
<name type="personal">
<namePart type="given">Minghao</namePart>
<namePart type="family">Zhu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junli</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chungang</namePart>
<namePart type="family">Yan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2022-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies</title>
</titleInfo>
<name type="personal">
<namePart type="given">Marine</namePart>
<namePart type="family">Carpuat</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marie-Catherine</namePart>
<namePart type="family">de Marneffe</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ivan</namePart>
<namePart type="given">Vladimir</namePart>
<namePart type="family">Meza Ruiz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Seattle, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Variational Autoencoder (VAE) is an effective framework to model the interdependency for non-autoregressive neural machine translation (NAT). One of the prominent VAE-based NAT frameworks, LaNMT, achieves great improvements to vanilla models, but still suffers from two main issues which lower down the translation quality: (1) mismatch between training and inference circumstances and (2) inadequacy of latent representations. In this work, we target on addressing these issues by proposing posterior consistency regularization. Specifically, we first perform stochastic data augmentation on the input samples to better adapt the model for inference circumstance, and then conduct consistency training on posterior latent variables to construct a more robust latent representations without any expansion on latent size. Experiments on En\textless-\textgreaterDe and En\textless-\textgreaterRo benchmarks confirm the effectiveness of our methods with about 1.5/0.7 and 0.8/0.3 BLEU points improvement to the baseline model with about 12.6\times faster than autoregressive Transformer.</abstract>
<identifier type="citekey">zhu-etal-2022-non</identifier>
<identifier type="doi">10.18653/v1/2022.naacl-main.45</identifier>
<location>
<url>https://aclanthology.org/2022.naacl-main.45</url>
</location>
<part>
<date>2022-07</date>
<extent unit="page">
<start>607</start>
<end>617</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Non-Autoregressive Neural Machine Translation with Consistency Regularization Optimized Variational Framework
%A Zhu, Minghao
%A Wang, Junli
%A Yan, Chungang
%Y Carpuat, Marine
%Y de Marneffe, Marie-Catherine
%Y Meza Ruiz, Ivan Vladimir
%S Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
%D 2022
%8 July
%I Association for Computational Linguistics
%C Seattle, United States
%F zhu-etal-2022-non
%X Variational Autoencoder (VAE) is an effective framework to model the interdependency for non-autoregressive neural machine translation (NAT). One of the prominent VAE-based NAT frameworks, LaNMT, achieves great improvements to vanilla models, but still suffers from two main issues which lower down the translation quality: (1) mismatch between training and inference circumstances and (2) inadequacy of latent representations. In this work, we target on addressing these issues by proposing posterior consistency regularization. Specifically, we first perform stochastic data augmentation on the input samples to better adapt the model for inference circumstance, and then conduct consistency training on posterior latent variables to construct a more robust latent representations without any expansion on latent size. Experiments on En\textless-\textgreaterDe and En\textless-\textgreaterRo benchmarks confirm the effectiveness of our methods with about 1.5/0.7 and 0.8/0.3 BLEU points improvement to the baseline model with about 12.6\times faster than autoregressive Transformer.
%R 10.18653/v1/2022.naacl-main.45
%U https://aclanthology.org/2022.naacl-main.45
%U https://doi.org/10.18653/v1/2022.naacl-main.45
%P 607-617
Markdown (Informal)
[Non-Autoregressive Neural Machine Translation with Consistency Regularization Optimized Variational Framework](https://aclanthology.org/2022.naacl-main.45) (Zhu et al., NAACL 2022)
ACL