Tower v2: Unbabel-IST 2024 Submission for the General MT Shared Task
Lists
Only you see this
Abstract
In this work, we present Tower v2, an improved iteration of the state-of-the-art open-weight Tower models, and the backbone of our submission to the WMT24 General Translation shared task. Tower v2 introduces key improvements including expanded language coverage, enhanced data quality, and increased model capacity up to 70B parameters. Our final submission combines these advancements with quality-aware decoding strategies, selecting translations based on multiple translation quality signals. The resulting system demonstrates significant improvement over previous versions, outperforming closed commercial systems like GPT-4o, Claude 3.5, and DeepL even at a smaller 7B scale.
BibTeX
@inproceedings{rei2024tower,
author = {Ricardo Rei and José Pombal and Nuno Miguel Guerreiro and João Alves and Pedro Henrique Martins and Patrick Fernandes and Helena Wu and Tânia Vaz and Duarte M. Alves and M. Amin Farajian and Sweta Agrawal and António Farinhas and José Guilherme Camargo de Souza and André F. T. Martins},
title = {{Tower v2: Unbabel-IST 2024 Submission for the General MT Shared Task}},
booktitle = {WMT},
year = {2024},
doi = {10.18653/v1/2024.wmt-1.12},
}Loading related content...