Adapters: A Unified Library for Parameter-Efficient and Modular Transfer Learning

TL;DR

Adapters is an open-source library that unifies various parameter-efficient transfer learning methods for large language models, enabling flexible, modular, and efficient adaptation across NLP tasks.

Contribution

The paper introduces a comprehensive library integrating 10 adapter methods into a single interface, facilitating modular transfer learning in NLP.

Findings

Library achieves comparable performance to full fine-tuning

Enables complex adapter compositions for flexible model adaptation

Promotes efficient transfer learning with modular design

Abstract

We introduce Adapters, an open-source library that unifies parameter-efficient and modular transfer learning in large language models. By integrating 10 diverse adapter methods into a unified interface, Adapters offers ease of use and flexible configuration. Our library allows researchers and practitioners to leverage adapter modularity through composition blocks, enabling the design of complex adapter setups. We demonstrate the library's efficacy by evaluating its performance against full fine-tuning on various NLP tasks. Adapters provides a powerful tool for addressing the challenges of conventional fine-tuning paradigms and promoting more efficient and modular transfer learning. The library is available via https://adapterhub.ml/adapters.