BERTGen: Multi-task Generation through BERT
Faidon Mitzalis, Ozan Caglayan, Pranava Madhyastha, Lucia Specia
Abstract
We present BERTGen, a novel, generative, decoder-only model which extends BERT by fusing multimodal and multilingual pre-trained models VL-BERT and M-BERT, respectively. BERTGen is auto-regressively trained for language generation tasks, namely image captioning, machine translation and multimodal machine translation, under a multi-task setting. With a comprehensive set of evaluations, we show that BERTGen outperforms many strong baselines across the tasks explored. We also show BERTGen’s ability for zero-shot language generation, where it exhibits competitive performance to supervised counterparts. Finally, we conduct ablation studies which demonstrate that BERTGen substantially benefits from multi-tasking and effectively transfers relevant inductive biases from the pre-trained models.
BibTeX
@inproceedings{mitzalis-etal-2021-bertgen,
title = "{BERTG}en: Multi-task Generation through {BERT}",
author = "Mitzalis, Faidon and
Caglayan, Ozan and
Madhyastha, Pranava and
Specia, Lucia",
editor = "Zong, Chengqing and
Xia, Fei and
Li, Wenjie and
Navigli, Roberto",
booktitle = "Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)",
month = aug,
year = "2021",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.acl-long.503/",
doi = "10.18653/v1/2021.acl-long.503",
pages = "6440--6455"
}