A Deep Generative Modeling Architecture for Designing Lattice-Constrained Materials

This dataset supports the article: contains all Crystallographic Information Files (CIF) and Quantum Espresso (QE) geometry optimization output files for the 124 perovskite materials that were newly designed using the Lattice-Constrained Materials Generative Model (LCMGM). CIF files were written using the Python Materials Genomics (Pymatgen) software package. For geometrical optimization, QE performs spin-polarized plane-wave calculations using the Perdew-Burke-Ernzerhof (PBE) pseudopotential class. The preprocessed dataset used for deep learning, relevant source codes for developing the LCMGM model, and Crystallographic Information Files (CIF) of newly designed materials are equally made available on GitHub: http://github.com/chenebuah/LCMGM