paper-with-me

Papers

Learning a Better Initialization for Soft Prompts via Meta-Learning

2022-05-25 · Yukun Huang, Kun Qian, Zhou Yu

Prompt tuning (PT) is an effective approach to adapting pre-trained language models to downstream tasks. Without a good initialization, prompt tuning doesn't perform well under few-shot settings. So pre-trained prompt tuning (PPT) is proposed to initialize prompts by leveraging pre-training data. We propose MetaPT (Meta-learned Prompt Tuning) to further improve PPT's initialization by considering latent structure within the pre-training data. Specifically, we introduce the structure by first clustering pre-training data into different auxiliary tasks with unsupervised methods. Then we use these tasks to pre-train prompts with a meta-learning algorithm. Such a process can make prompts learn a better initialization by discovering commonalities among these auxiliary tasks. We evaluate our method on seven downstream tasks. Our MetaPT achieves better and more stable performance than the state-of-the-art method.

📄 PDF Abstract BibTeX arXiv:2205.12471

Code (0)

등록된 구현이 없습니다.

Tasks

Meta-Learning

Similar Papers 제목 키워드 기반

MetaPrompting: Learning to Learn Better Prompts

2022-09-23 · COLING 2022 10 · Yutai Hou, Hongyuan Dong, Xinghao Wang, Bohan Li 외

Prompting method is regarded as one of the crucial progress for few-shot nature language processing. Recent research on prompting moves from discrete tokens based ``hard prompts'' to continuous ``soft prompts'', which em…

Meta-Learning

MetaPrompting: Learning to Learn Better Prompts

2021-11-16 · ACL ARR November 2021 11 · Anonymous

Prompting method is regarded as one of the crucial progress for few-shot nature language processing. Recent research on prompting moves from discrete tokens based "hard prompts" to continuous "soft prompts", which employ…

Meta-Learning

Effective Structured Prompting by Meta-Learning and Representative Verbalizer

2023-06-01 · Weisen Jiang, Yu Zhang, James T. Kwok

Prompt tuning for pre-trained masked language models (MLM) has shown promising performance in natural language processing tasks with few labeled examples. It tunes a prompt for the downstream task, and a verbalizer is us…

Meta-Learning

Self-supervised Meta-Prompt Learning with Meta-Gradient Regularization for Few-shot Generalization

2023-03-22 · Kaihang Pan, Juncheng Li, Hongye Song, Jun Lin 외

Prompt tuning is a parameter-efficient method, which learns soft prompts and conditions frozen language models to perform specific downstream tasks. Though effective, prompt tuning under few-shot settings on the one hand…

Domain GeneralizationFew-Shot LearningMeta-LearningPrompt Learning

Gradient-Regulated Meta-Prompt Learning for Generalizable Vision-Language Models

2023-03-12 · ICCV 2023 1 · Juncheng Li, Minghe Gao, Longhui Wei, Siliang Tang 외

Prompt tuning, a recently emerging paradigm, enables the powerful vision-language pre-training models to adapt to downstream tasks in a parameter -- and data -- efficient way, by learning the ``soft prompts'' to conditio…

Domain GeneralizationFew-Shot LearningMeta-LearningPrompt Learning+1