Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks

Suchin Gururangan, Ana Marasović, Swabha Swayamdipta, Kyle Lo, Iz Beltagy, Doug Downey, Noah A. Smith


Abstract
Language models pretrained on text from a wide variety of sources form the foundation of today’s NLP. In light of the success of these broad-coverage models, we investigate whether it is still helpful to tailor a pretrained model to the domain of a target task. We present a study across four domains (biomedical and computer science publications, news, and reviews) and eight classification tasks, showing that a second phase of pretraining in-domain (domain-adaptive pretraining) leads to performance gains, under both high- and low-resource settings. Moreover, adapting to the task’s unlabeled data (task-adaptive pretraining) improves performance even after domain-adaptive pretraining. Finally, we show that adapting to a task corpus augmented using simple data selection strategies is an effective alternative, especially when resources for domain-adaptive pretraining might be unavailable. Overall, we consistently find that multi-phase adaptive pretraining offers large gains in task performance.
Anthology ID:
2020.acl-main.740
Volume:
Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics
Month:
July
Year:
2020
Address:
Online
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
8342–8360
Language:
URL:
https://aclanthology.org/2020.acl-main.740
DOI:
10.18653/v1/2020.acl-main.740
Award:
 Honorable Mention for Best Overall Paper
Bibkey:
Copy Citation:
PDF:
https://aclanthology.org/2020.acl-main.740.pdf
Video:
 http://slideslive.com/38929123