Learning Earthquake Ground Motions via Conditional Generative Modeling

v1.0.0 - Release Notes Initial Release This release provides the official implementation of CGM-GM, a conditional generative modeling framework for learning physics-informed earthquake ground motions. Key features: Conditional generative model for spatially heterogeneous ground motion modeling Physics-aware learning using earthquake and sensor metadata Training and evaluation pipeline with configurable hyperparameters Reproducible experiments corresponding to the paper results Paper: https://arxiv.org/abs/2407.15089