A Lipschitz Bandits Approach for Continuous Hyperparameter Optimization
Abstract
One of the most critical problems in machine learning is HyperParameter Optimization (HPO), since choice of hyperparameters has a significant impact on final model performance. Although there are many HPO algorithms, they either have no theoretical guarantees or require strong assumptions. To this end, we introduce BLiE -- a Lipschitz-bandit-based algorithm for HPO that only assumes Lipschitz continuity of the objective function. BLiE exploits the landscape of the objective function to adaptively search over the hyperparameter space. Theoretically, we show that BLiE finds an -optimal hyperparameter with total budgets, where and are problem intrinsic; BLiE is highly parallelizable. Empirically, we demonstrate that BLiE outperforms the state-of-the-art HPO algorithms on benchmark tasks. We also apply BLiE to search for noise schedule of diffusion models. Comparison with the default schedule shows that BLiE schedule greatly improves the sampling speed.
Keywords
Cite
@article{arxiv.2302.01539,
title = {A Lipschitz Bandits Approach for Continuous Hyperparameter Optimization},
author = {Yasong Feng and Weijian Luo and Yimin Huang and Tianyu Wang},
journal= {arXiv preprint arXiv:2302.01539},
year = {2023}
}
Comments
Some preliminaries and backgrounds are drawn from arXiv:2110.09722 by the first author and the last author, and their coauthor Z. Huang