ATTEMPT: Parameter-Efficient Multi-task Tuning via Attentional Mixtures of Soft Prompts

Akari Asai, Mohammadreza Salehi, Matthew Peters, Hannaneh Hajishirzi


Abstract
This work introduces a new multi-task, parameter-efficient language model (LM) tuning method that learns to transfer knowledge across different tasks via a mixture of soft prompts—small prefix embedding vectors pre-trained for different tasks. Our method, called ATTEMPT (ATTEntional Mixtures of Prompt Tuning), obtains source prompts as encodings of large-scale source tasks into a small number of parameters and trains an attention module to interpolate the source prompts and a newly initialized target prompt for every instance in the target task. During training, only the target task prompt and the attention weights, which are shared between tasks in multi-task training, are updated, while the original LM and source prompts are intact. ATTEMPT is highly parameter-efficient (e.g., updates 2,300 times fewer parameters than full fine-tuning), while it overcomes instability of prompt tuning and achieves high task performance using learned knowledge from high-resource tasks. Moreover, it is modular using pre-trained soft prompts, and can flexibly add or remove source prompts for effective knowledge transfer. Our experimental results across 21 diverse NLP datasets show that ATTEMPT significantly outperforms prompt tuning and outperforms or matches fully fine-tuned or other parameter-efficient tuning approaches that use 10 times more parameters. Finally, ATTEMPT outperforms previous work in few-shot learning settings.
Anthology ID:
2022.emnlp-main.446
Volume:
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Editors:
Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
6655–6672
Language:
URL:
https://aclanthology.org/2022.emnlp-main.446
DOI:
10.18653/v1/2022.emnlp-main.446
Bibkey:
Cite (ACL):
Akari Asai, Mohammadreza Salehi, Matthew Peters, and Hannaneh Hajishirzi. 2022. ATTEMPT: Parameter-Efficient Multi-task Tuning via Attentional Mixtures of Soft Prompts. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 6655–6672, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
ATTEMPT: Parameter-Efficient Multi-task Tuning via Attentional Mixtures of Soft Prompts (Asai et al., EMNLP 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.emnlp-main.446.pdf