Files for GCMC Adlayers and MACE Fine-tuning/Inference

This dataset contains the output XYZ structure files from Grand Canonical Monte Carlo (GCMC) sampling of atomic S and O adsorption on Ni(111), calculated using the SuSMoST software and parameterised using density functional theory using the FHI-aims code. Also included are files for the fine-tuning and inference of a MACE machine learned interatomic potential. The model is trained using a diverse dataset of FHI-aims single point evaluations of energies and forces. The model is tested for geometry optimisation calculations of six GCMC-predicted adlayer structures.