Prompt Tuning Strikes Back: Customizing Foundation Models with Low-Rank Prompt Adaptation

TL;DR

This paper introduces Low-Rank Prompt Adaptation (LoPA), a parameter-efficient prompt tuning method that rivals state-of-the-art PEFT techniques and full fine-tuning, while avoiding the scalability issues of traditional adapters.

Contribution

LoPA is a novel prompt tuning approach using low-rank decomposition to enhance parameter efficiency and performance across diverse tasks and models.

Findings

LoPA matches the performance of state-of-the-art PEFT methods.

LoPA outperforms traditional prompt tuning in efficiency.

LoPA is effective across multiple NLP and code tasks.

Abstract

Parameter-Efficient Fine-Tuning (PEFT) has become the standard for customising Foundation Models (FMs) to user-specific downstream tasks. However, typical PEFT methods require storing multiple task-specific adapters, creating scalability issues as these adapters must be housed and run at the FM server. Traditional prompt tuning offers a potential solution by customising them through task-specific input prefixes, but it under-performs compared to other PEFT methods like LoRA. To address this gap, we propose Low-Rank Prompt Adaptation (LoPA), a prompt-tuning-based approach that performs on par with state-of-the-art PEFT methods and full fine-tuning while being more parameter-efficient and not requiring a server-based adapter. LoPA generates soft prompts by balancing between sharing task-specific information across instances and customization for each instance. It uses a low-rank decomposition of the soft-prompt component encoded for each instance to achieve parameter efficiency. We provide a comprehensive evaluation on multiple natural language understanding and code generation and understanding tasks across a wide range of foundation models with varying sizes.