The mergers of binary compact objects are of central interest to several areas of astrophysics, including as the progenitors of short gamma-ray bursts (GRBs). Dozens of GRBs have been confidently associated with rapidly-decaying optical transients (``afterglows'') and more recently with late-time emission, such as from a kilonovae. Traditional methods for modeling these phenomena are computationally expensive for comparison and parameter inference because of the high-dimension parameter space. We propose using Simulation-Based Inference (SBI) as a fast and scalable alternative to conventional likelihood-based approaches. Preliminary results, using SBI to fit multi-wavelength light curves of GRB\,230307A across different emission models, highlight its efficiency in managing high-dimensional parameter space and show that SBI yields posterior distributions consistent with those from likelihood-based methods, while significantly reducing computational time.