AdapterGNN: Parameter-Efficient Fine-Tuning Improves Generalization in GNNs
Authors:
Preprint
Abstract
Fine-tuning pre-trained models has recently yielded remarkable performance gains in graph neural networks (GNNs). In addition to pre-training techniques, inspired by the latest work in the natural language fields, more recent work has shifted towards applying effective fine-tuning approaches, such as parameter-efficient fine-tuning (PEFT). However, given the substantial differences between GNNs and transformer-based models, applying such approaches directly to GNNs proved to be less effective. In this paper, we present a comprehensive comparison of PEFT techniques for GNNs and propose a novel PEFT method specifically designed for GNNs, called AdapterGNN. AdapterGNN preserves the knowledge of the large pre-trained model and leverages highly expressive adapters for GNNs, which can adapt to downstream tasks effectively with only a few parameters, while also improving the model's generalization ability. Extensive experiments show that AdapterGNN achieves higher performance than other PEFT methods and is the only one consistently surpassing full fine-tuning (outperforming it by 1.6% and 5.7% in the chemistry and biology domains respectively, with only 5% and 4% of its parameters tuned) with lower generalization gaps. Moreover, we empirically show that a larger GNN model can have a worse generalization ability, which differs from the trend observed in large transformer-based models. Building upon this, we provide a theoretical justification for PEFT can improve generalization of GNNs by applying generalization bounds. Our code is available at https://github.com/Lucius-lsr/AdapterGNN.Keywords
All Related Versions
This publication has 15 references indexed in Scilit:
- GraphPrompt: Unifying Pre-Training and Downstream Tasks for Graph Neural Networks2023
- Advanced graph and sequence neural networks for molecular property prediction and drug discoveryBioinformatics, 2022
- Deep double descent: where bigger models and more data hurt*Journal of Statistical Mechanics: Theory and Experiment, 2021
- Understanding deep learning (still) requires rethinking generalizationCommunications of the ACM, 2021
- Pre-trained models: Past, present and futureAI Open, 2021
- Prefix-Tuning: Optimizing Continuous Prompts for Generation2021
- A Comprehensive Survey on Graph Neural NetworksIEEE Transactions on Neural Networks and Learning Systems, 2020
- Reconciling modern machine-learning practice and the classical bias–variance trade-offProceedings of the National Academy of Sciences of the United States of America, 2019
- ZINC 15 – Ligand Discovery for EveryoneJournal of Chemical Information and Modeling, 2015
- Over-Fitting and Model Tuning2013