Model driver for the Generalized Embedded Atom Method (GEAM) potential v001

This model driver is a C++ implementation of the Generalized Embedded Atom Method (GEAM) developed by Amit Samanta and coworkers, originally implemented in LAMMPS format. GEAM is a data-driven interatomic potential for multi-element systems. It incorporates embedding energy and two-, three-, and nonlocal many-body effects through Gaussian basis functions, providing sufficient flexibility to model different atomic environments. The parameters of the potential are optimized using training data generated through ab initio density functional theory (DFT) calculations. The GEAM potential was originally developed to more accurately model BCC metals than the EAM, with the accuracy in predicting physical properties of machine learning potentials such as GAP, while being less expensive and requiring fewer training structures. A parameterization (KIM model) utilizing the GEAM driver must provide: 1) a file containing the element symbols, 2) a file containing the LAMMPS-formatted GEAM pair_style parameters, and 3) a file containing the optimized parameters from the training data.