IDPG: An Instance-Dependent Prompt Generation Method

Autor:	Wu, Zhuofeng, Wang, Sinong, Gu, Jiatao, Hou, Rui, Dong, Yuxiao, Vydiswaran, V. G. Vinod, Ma, Hao
Rok vydání:	2022
Předmět:	Computer Science - Computation and Language Computer Science - Machine Learning
Druh dokumentu:	Working Paper
Popis:	Prompt tuning is a new, efficient NLP transfer learning paradigm that adds a task-specific prompt in each input instance during the model training stage. It freezes the pre-trained language model and only optimizes a few task-specific prompts. In this paper, we propose a conditional prompt generation method to generate prompts for each input instance, referred to as the Instance-Dependent Prompt Generation (IDPG). Unlike traditional prompt tuning methods that use a fixed prompt, IDPG introduces a lightweight and trainable component to generate prompts based on each input sentence. Extensive experiments on ten natural language understanding (NLU) tasks show that the proposed strategy consistently outperforms various prompt tuning baselines and is on par with other efficient transfer learning methods such as Compacter while tuning far fewer model parameters. Comment: To appear at the NAACL 2022 main conference
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2204.04497 Zobrazit plný text záznamu View this record from Arxiv