Supporting data of "DeepH-pack: A general-purpose neural network package for deep-learning electronic structure calculations"

This database contains the raw data underlying the analysis presented in the "DeepH-pack: A general-purpose neural network package for deep-learning electronic structure calculations" The benchmark_data.tgz contains the necessary data to reproduce the results of Table 1 and Table 2 in the article. The test set of seven datasets are provided: H2O, MoS2, Bi2Se3, Graphite, Cu, SiC, and Carbon. Three types of trained neural network models are provided: normal, accurate, and accurate_uniformed. Detailed information can be found in the README within the tgz file. For more information about the example applications, one can be refer to https://drive.google.com/drive/folders/116Xpgcbc-kHR4LK089-h9WrdmIqGiFsQ