Non-Autoregressive Translation by Learning Target Categorical Codes

Yu Bao, Shujian Huang, Tong Xiao, Dongqi Wang, Xinyu Dai, Jiajun Chen


Abstract
Non-autoregressive Transformer is a promising text generation model. However, current non-autoregressive models still fall behind their autoregressive counterparts in translation quality. We attribute this accuracy gap to the lack of dependency modeling among decoder inputs. In this paper, we propose CNAT, which learns implicitly categorical codes as latent variables into the non-autoregressive decoding. The interaction among these categorical codes remedies the missing dependencies and improves the model capacity. Experiment results show that our model achieves comparable or better performance in machine translation tasks than several strong baselines.
Anthology ID:
2021.naacl-main.458
Volume:
Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
Month:
June
Year:
2021
Address:
Online
Editors:
Kristina Toutanova, Anna Rumshisky, Luke Zettlemoyer, Dilek Hakkani-Tur, Iz Beltagy, Steven Bethard, Ryan Cotterell, Tanmoy Chakraborty, Yichao Zhou
Venue:
NAACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
5749–5759
Language:
URL:
https://aclanthology.org/2021.naacl-main.458
DOI:
10.18653/v1/2021.naacl-main.458
Bibkey:
Cite (ACL):
Yu Bao, Shujian Huang, Tong Xiao, Dongqi Wang, Xinyu Dai, and Jiajun Chen. 2021. Non-Autoregressive Translation by Learning Target Categorical Codes. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 5749–5759, Online. Association for Computational Linguistics.
Cite (Informal):
Non-Autoregressive Translation by Learning Target Categorical Codes (Bao et al., NAACL 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.naacl-main.458.pdf
Video:
 https://aclanthology.org/2021.naacl-main.458.mp4
Code
 baoy-nlp/CNAT
Data
WMT 2014