Knowledge Inheritance for Pre-trained Language Models
Yujia Qin, Yankai Lin, Jing Yi, Jiajie Zhang, Xu Han, Zhengyan Zhang, Yusheng Su, Zhiyuan Liu
Abstract
Recent explorations of large-scale pre-trained language models (PLMs) have revealed the power of PLMs with huge amounts of parameters, setting off a wave of training ever-larger PLMs. However, it requires tremendous computational resources to train a large-scale PLM, which may be practically unaffordable. In addition, existing large-scale PLMs are mainly trained from scratch individually, ignoring that many well-trained PLMs are available. To this end, we explore the question how could existing PLMs benefit training large-scale PLMs in future. Specifically, we introduce a pre-training framework named “knowledge inheritance” (KI) and explore how could knowledge distillation serve as auxiliary supervision during pre-training to efficiently learn larger PLMs. Experimental results demonstrate the superiority of KI in training efficiency. We also conduct empirical analyses to explore the effects of teacher PLMs’ pre-training settings, including model architecture, pre-training data, etc. Finally, we show that KI could be applied to domain adaptation and knowledge transfer.
BibTeX
@inproceedings{qin-etal-2022-knowledge,
title = "Knowledge Inheritance for Pre-trained Language Models",
author = "Qin, Yujia and
Lin, Yankai and
Yi, Jing and
Zhang, Jiajie and
Han, Xu and
Zhang, Zhengyan and
Su, Yusheng and
Liu, Zhiyuan and
Li, Peng and
Sun, Maosong and
Zhou, Jie",
editor = "Carpuat, Marine and
de Marneffe, Marie-Catherine and
Meza Ruiz, Ivan Vladimir",
booktitle = "Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies",
month = jul,
year = "2022",
address = "Seattle, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.naacl-main.288/",
doi = "10.18653/v1/2022.naacl-main.288",
pages = "3921--3937"
}