AdapterFusion: Non-Destructive Task Composition for Transfer Learning

Jonas Pfeiffer, Aishwarya Kamath, Andreas Rücklé, Kyunghyun Cho, Iryna Gurevych


Abstract
Sequential fine-tuning and multi-task learning are methods aiming to incorporate knowledge from multiple tasks; however, they suffer from catastrophic forgetting and difficulties in dataset balancing. To address these shortcomings, we propose AdapterFusion, a new two stage learning algorithm that leverages knowledge from multiple tasks. First, in the knowledge extraction stage we learn task specific parameters called adapters, that encapsulate the task-specific information. We then combine the adapters in a separate knowledge composition step. We show that by separating the two stages, i.e., knowledge extraction and knowledge composition, the classifier can effectively exploit the representations learned from multiple tasks in a non-destructive manner. We empirically evaluate AdapterFusion on 16 diverse NLU tasks, and find that it effectively combines various types of knowledge at different layers of the model. We show that our approach outperforms traditional strategies such as full fine-tuning as well as multi-task learning. Our code and adapters are available at AdapterHub.ml.
Anthology ID:
2021.eacl-main.39
Volume:
Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume
Month:
April
Year:
2021
Address:
Online
Venue:
EACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
487–503
Language:
URL:
https://aclanthology.org/2021.eacl-main.39
DOI:
Bibkey:
Copy Citation:
PDF:
https://aclanthology.org/2021.eacl-main.39.pdf