Dataset: Benchmarking Universal Machine Learning Interatomic Potentials for Real-Time Analysis of Inelastic Neutron Scattering Data

The dataset contains the unit cell structures (POSCAR-unitcell), force sets calculated with DFT, input files for Phonopy calculations and optional BORN files of 4869 materials. The data is used to generate the phonon properties with selected universal machine learning interatomic potentials. The phonon properties can be generated with the Python script provided in the GitHub repository (https://github.com/maplewen4/phonon_uMLIP).