Parameter-Efficient Tuning of Large Convolutional Models

arXiv:2403.00269v1 Announce Type: cross
Abstract: To address the high computational and parameter complexity associated with fine-tuning large pre-trained models, researchers have developed parameter-efficient methods, where only partial parameters are updated for downstream tasks. However, these works often overlook the distinct properties of convolutional kernels, which still remain essential elements in many large models, such as Stable Diffusion. In this study, we first introduce filter subspace by decomposing convolutional kernels within each network layer over a small set of filter subspace …

abstract arxiv complexity computational cs.cv cs.lg fine-tuning large models parameters pre-trained models researchers tasks type