The AISP-SJTU Translation System for WMT 2022

Guangfeng Liu, Qinpei Zhu, Xingyu Chen, Renjie Feng, Jianxin Ren, Renshou Wu, Qingliang Miao, Rui Wang, Kai Yu


Abstract
This paper describes AISP-SJTU’s participation in WMT 2022 shared general MT task. In this shared task, we participated in four translation directions: English-Chinese, Chinese-English, English-Japanese and Japanese-English. Our systems are based on the Transformer architecture with several novel and effective variants, including network depth and internal structure. In our experiments, we employ data filtering, large-scale back-translation, knowledge distillation, forward-translation, iterative in-domain knowledge finetune and model ensemble. The constrained systems achieve 48.8, 29.7, 39.3 and 22.0 case-sensitive BLEU scores on EN-ZH, ZH-EN, EN-JA and JA-EN, respectively.
Anthology ID:
2022.wmt-1.24
Volume:
Proceedings of the Seventh Conference on Machine Translation (WMT)
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates (Hybrid)
Venue:
WMT
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
310–317
Language:
URL:
https://aclanthology.org/2022.wmt-1.24
DOI:
Bibkey:
Cite (ACL):
Guangfeng Liu, Qinpei Zhu, Xingyu Chen, Renjie Feng, Jianxin Ren, Renshou Wu, Qingliang Miao, Rui Wang, and Kai Yu. 2022. The AISP-SJTU Translation System for WMT 2022. In Proceedings of the Seventh Conference on Machine Translation (WMT), pages 310–317, Abu Dhabi, United Arab Emirates (Hybrid). Association for Computational Linguistics.
Cite (Informal):
The AISP-SJTU Translation System for WMT 2022 (Liu et al., WMT 2022)
Copy Citation:
PDF:
https://preview.aclanthology.org/ingestion-script-update/2022.wmt-1.24.pdf