Learning Task Sampling Policy for Multitask Learning

Dhanasekar Sundararaman, Henry Tsai, Kuang-Huei Lee, Iulia Turc, Lawrence Carin


Abstract
It has been shown that training multi-task models with auxiliary tasks can improve the target task quality through cross-task transfer. However, the importance of each auxiliary task to the primary task is likely not known a priori. While the importance weights of auxiliary tasks can be manually tuned, it becomes practically infeasible with the number of tasks scaling up. To address this, we propose a search method that automatically assigns importance weights. We formulate it as a reinforcement learning problem and learn a task sampling schedule based on the evaluation accuracy of the multi-task model. Our empirical evaluation on XNLI and GLUE shows that our method outperforms uniform sampling and the corresponding single-task baseline.
Anthology ID:
2021.findings-emnlp.375
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2021
Month:
November
Year:
2021
Address:
Punta Cana, Dominican Republic
Editors:
Marie-Francine Moens, Xuanjing Huang, Lucia Specia, Scott Wen-tau Yih
Venue:
Findings
SIG:
SIGDAT
Publisher:
Association for Computational Linguistics
Note:
Pages:
4410–4415
Language:
URL:
https://aclanthology.org/2021.findings-emnlp.375
DOI:
10.18653/v1/2021.findings-emnlp.375
Bibkey:
Cite (ACL):
Dhanasekar Sundararaman, Henry Tsai, Kuang-Huei Lee, Iulia Turc, and Lawrence Carin. 2021. Learning Task Sampling Policy for Multitask Learning. In Findings of the Association for Computational Linguistics: EMNLP 2021, pages 4410–4415, Punta Cana, Dominican Republic. Association for Computational Linguistics.
Cite (Informal):
Learning Task Sampling Policy for Multitask Learning (Sundararaman et al., Findings 2021)
Copy Citation:
PDF:
https://preview.aclanthology.org/naacl24-info/2021.findings-emnlp.375.pdf
Video:
 https://preview.aclanthology.org/naacl24-info/2021.findings-emnlp.375.mp4
Data
GLUEMultiNLIQNLI