Skip to content
/ ITST Public

Code for EMNLP 2022 main conference paper "Information-Transport-based Policy for Simultaneous Translation"

License

Notifications You must be signed in to change notification settings

ictnlp/ITST

Folders and files

NameName
Last commit message
Last commit date

Latest commit

b6b0d9c · Nov 3, 2022

History

7 Commits
Nov 3, 2022
Oct 22, 2022
Oct 22, 2022
Oct 25, 2022
Oct 22, 2022
Oct 22, 2022
Oct 22, 2022
Oct 22, 2022
Nov 3, 2022
Oct 22, 2022
Nov 3, 2022
Nov 3, 2022
Nov 3, 2022
Nov 3, 2022
Oct 22, 2022
Oct 22, 2022
Oct 22, 2022
Oct 22, 2022
Oct 22, 2022
Oct 22, 2022

Repository files navigation

Information-Transport-based Policy for Simultaneous Translation

Source code for our EMNLP 2022 main conference paper "Information-Transport-based Policy for Simultaneous Translation"

fig

Our method is implemented based on the open-source toolkit Fairseq.

Requirements and Installation

  • Python version = 3.8

  • PyTorch version = 1.7.1

  • Install requirements:

    git clone https://github.com/ictnlp/ITST.git
    cd ITST
    pip install -r requirements.txt
    pip install --editable ./

Quick Start

Information-transport-based simultaneous translation (ITST) achieves good results on both text-to-text simultaneous translation and speech-to-text simultaneous translation (a.k.a., streaming speech translation). Detailed introductions refer to:

All example shell scripts refer to shell_scripts/.

Citation

If you have any questions, feel free to contact me with: zhangshaolei20z@ict.ac.cn.

If this repository is useful for you, please cite as:

@inproceedings{ITST,
    title = "Information-Transport-based Policy for Simultaneous Translation",
    author = "Zhang, Shaolei  and
      Feng, Yang",
    booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing",
    month = dec,
    year = "2022",
    address = "Online and Abu Dhabi",
    publisher = "Association for Computational Linguistics",
    url="https://arxiv.org/pdf/2210.12357.pdf",
}