Exploring Sparsity for Parameter Efficient Fine Tuning Using Wavelets

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Bilican, Ahmet, Yılmaz, M. Akın, Tekalp, A. Murat, Cinbiş, R. Gökberk
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918045249699840
author Bilican, Ahmet
Yılmaz, M. Akın
Tekalp, A. Murat
Cinbiş, R. Gökberk
author_facet Bilican, Ahmet
Yılmaz, M. Akın
Tekalp, A. Murat
Cinbiş, R. Gökberk
contents Efficiently adapting large foundation models is critical, especially with tight compute and memory budgets. Parameter-Efficient Fine-Tuning (PEFT) methods such as LoRA offer limited granularity and effectiveness in few-parameter regimes. We propose Wavelet Fine-Tuning (WaveFT), a novel PEFT method that learns highly sparse updates in the wavelet domain of residual matrices. WaveFT allows precise control of trainable parameters, offering fine-grained capacity adjustment and excelling with remarkably low parameter count, potentially far fewer than LoRA's minimum, ideal for extreme parameter-efficient scenarios. Evaluated on personalized text-to-image generation using Stable Diffusion XL as baseline, WaveFT significantly outperforms LoRA and other PEFT methods, especially at low parameter counts; achieving superior subject fidelity, prompt alignment, and image diversity.
format Preprint
id arxiv_https___arxiv_org_abs_2505_12532
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Exploring Sparsity for Parameter Efficient Fine Tuning Using Wavelets
Bilican, Ahmet
Yılmaz, M. Akın
Tekalp, A. Murat
Cinbiş, R. Gökberk
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Image and Video Processing
Signal Processing
Efficiently adapting large foundation models is critical, especially with tight compute and memory budgets. Parameter-Efficient Fine-Tuning (PEFT) methods such as LoRA offer limited granularity and effectiveness in few-parameter regimes. We propose Wavelet Fine-Tuning (WaveFT), a novel PEFT method that learns highly sparse updates in the wavelet domain of residual matrices. WaveFT allows precise control of trainable parameters, offering fine-grained capacity adjustment and excelling with remarkably low parameter count, potentially far fewer than LoRA's minimum, ideal for extreme parameter-efficient scenarios. Evaluated on personalized text-to-image generation using Stable Diffusion XL as baseline, WaveFT significantly outperforms LoRA and other PEFT methods, especially at low parameter counts; achieving superior subject fidelity, prompt alignment, and image diversity.
title Exploring Sparsity for Parameter Efficient Fine Tuning Using Wavelets
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Image and Video Processing
Signal Processing
url https://arxiv.org/abs/2505.12532