With the advent in Additive Manufacturing (AM) technologies and Computational Sciences, design algorithms such as Topology Optimization (TO) have garnered the interest of academia and industry. TO aims to generate optimum structures by maximizing the stiffness of the structure, given a set of geometric, loading and boundary conditions. However, these approaches are computationally expensive as it requires many iterations to converge to an optimum solution. The purpose of this work is to explore the effectiveness of deep generative models on a diverse range of topology optimization problems with varying design constraints, loading and boundary conditions. Specifically, four distinctive models were successfully developed, trained, and evaluated to generate rapid designs with comparable results to that of conventional algorithms. Our findings highlight the effectiveness of the novel design problem representation and proposed generative models in rapid topology optimization.