colabfit/GST_GAP_22_extended

GST GAP 22扩展数据集是一个用于训练机器学习原子间势能的数据集，它基于PBEsol功能计算得出。该数据集包含了在外部电场下模拟的新配置，并使用DFT进行了标注，添加到了原始的参考数据库GST-GAP-22中。GST-GAP-22包含了在GeTe-Sb2Te3（GST）伪二元化学成分线上的相变材料配置。这些数据被用于在真实设备条件下模拟一系列的锗-锑-碲成分。

The GST GAP 22 extended dataset is a training dataset for a machine learning interatomic potential, calculated using the PBEsol functional. It includes new configurations simulated under external electric fields, which are labeled with DFT and added to the original reference database GST-GAP-22. GST-GAP-22 contains configurations of phase-change materials on the quasi-binary GeTe-Sb2Te3 (GST) line of chemical compositions. The data is used for simulating a range of germanium-antimony-tellurium compositions under realistic device conditions.