English

Parameter-Efficient Fine-Tuning with Discrete Fourier Transform

Machine Learning 2024-05-07 v1 Artificial Intelligence Computation and Language

Abstract

Low-rank adaptation~(LoRA) has recently gained much interest in fine-tuning foundation models. It effectively reduces the number of trainable parameters by incorporating low-rank matrices AA and BB to represent the weight change, i.e., ΔW=BA\Delta W=BA. Despite LoRA's progress, it faces storage challenges when handling extensive customization adaptations or larger base models. In this work, we aim to further compress trainable parameters by enjoying the powerful expressiveness of the Fourier transform. Specifically, we introduce FourierFT, which treats ΔW\Delta W as a matrix in the spatial domain and learns only a small fraction of its spectral coefficients. With the trained spectral coefficients, we implement the inverse discrete Fourier transform to recover ΔW\Delta W. Empirically, our FourierFT method shows comparable or better performance with fewer parameters than LoRA on various tasks, including natural language understanding, natural language generation, instruction tuning, and image classification. For example, when performing instruction tuning on the LLaMA2-7B model, FourierFT surpasses LoRA with only 0.064M trainable parameters, compared to LoRA's 33.5M. Our code is released at \url{https://github.com/Chaos96/fourierft}.

Keywords

Cite

@article{arxiv.2405.03003,
  title  = {Parameter-Efficient Fine-Tuning with Discrete Fourier Transform},
  author = {Ziqi Gao and Qichao Wang and Aochuan Chen and Zijing Liu and Bingzhe Wu and Liang Chen and Jia Li},
  journal= {arXiv preprint arXiv:2405.03003},
  year   = {2024}
}

Comments

Accepted by ICML 2024

R2 v1 2026-06-28T16:17:17.987Z