Abstract
This paper presents ExPSO, a Python package designed to simplify parameter selection in deep learning models. ExPSO utilizes the Exponential Particle Swarm Optimization (ExPSO) method for global optimization problems, which has a superior ability to balance exploration and exploitation in search spaces. This package provides a user-friendly framework that promises to enhance the performance and evaluation of various deep learning algorithms through its exponential selection technique. In addition to its primary features, ExPSO is designed with extensibility in mind. It serves as a robust foundation for the development of innovative selection methodologies and can be easily adapted to incorporate other optimization algorithms and techniques. This flexibility ensures ExPSO remains relevant and useful as new advancements in the field of optimization and deep learning emerge.