From Mimicking to Integrating: Knowledge Integration for Pre-Trained Language Models

Lei Li, Yankai Lin, Xuancheng Ren, Guangxiang Zhao, Peng Li, Jie Zhou, Xu Sun


Abstract
Investigating better ways to reuse the released pre-trained language models (PLMs) can significantly reduce the computational cost and the potential environmental side-effects. This paper explores a novel PLM reuse paradigm, Knowledge Integration (KI). Without human annotations available, KI aims to merge the knowledge from different teacher-PLMs, each of which specializes in a different classification problem, into a versatile student model. To achieve this, we first derive the correlation between virtual golden supervision and teacher predictions. We then design a Model Uncertainty–aware Knowledge Integration (MUKI) framework to recover the golden supervision for the student. Specifically, MUKI adopts Monte-Carlo Dropout to estimate model uncertainty for the supervision integration. An instance-wise re-weighting mechanism based on the margin of uncertainty scores is further incorporated, to deal with the potential conflicting supervision from teachers.Experimental results demonstrate that MUKI achieves substantial improvements over baselines on benchmark datasets. Further analysis shows that MUKI can generalize well for merging teacher models with heterogeneous architectures, and even teachers major in cross-lingual datasets.
Anthology ID:
2022.findings-emnlp.477
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2022
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
6391–6402
Language:
URL:
https://aclanthology.org/2022.findings-emnlp.477
DOI:
Bibkey:
Cite (ACL):
Lei Li, Yankai Lin, Xuancheng Ren, Guangxiang Zhao, Peng Li, Jie Zhou, and Xu Sun. 2022. From Mimicking to Integrating: Knowledge Integration for Pre-Trained Language Models. In Findings of the Association for Computational Linguistics: EMNLP 2022, pages 6391–6402, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
From Mimicking to Integrating: Knowledge Integration for Pre-Trained Language Models (Li et al., Findings 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.findings-emnlp.477.pdf