ZhiJian: A Unifying and Rapidly Deployable Toolbox for Pre-trained Model Reuse

TL;DR

ZhiJian is a versatile, easy-to-use toolbox that unifies various methods of pre-trained model reuse, enabling efficient deployment and exploration of downstream tasks in machine learning.

Contribution

It introduces a novel paradigm that unifies diverse perspectives on model reuse, providing a comprehensive toolkit for target architecture construction, tuning, and inference with pre-trained models.

Findings

Supports multiple model reuse strategies within a single framework

Facilitates rapid deployment and exploration of downstream tasks

Accessible via an open-source PyTorch toolbox

Abstract

The rapid expansion of foundation pre-trained models and their fine-tuned counterparts has significantly contributed to the advancement of machine learning. Leveraging pre-trained models to extract knowledge and expedite learning in real-world tasks, known as "Model Reuse", has become crucial in various applications. Previous research focuses on reusing models within a certain aspect, including reusing model weights, structures, and hypothesis spaces. This paper introduces ZhiJian, a comprehensive and user-friendly toolbox for model reuse, utilizing the PyTorch backend. ZhiJian presents a novel paradigm that unifies diverse perspectives on model reuse, encompassing target architecture construction with PTM, tuning target model with PTM, and PTM-based inference. This empowers deep learning practitioners to explore downstream tasks and identify the complementary advantages among different methods. ZhiJian is readily accessible at https://github.com/zhangyikaii/lamda-zhijian facilitating seamless utilization of pre-trained models and streamlining the model reuse process for researchers and developers.