colabfit/a-AlOx_JCP_2020
收藏Hugging Face2025-10-17 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/colabfit/a-AlOx_JCP_2020
下载链接
链接失效反馈官方服务：
资源简介：
a-AlOx JCP 2020数据集用于训练无定形氧化铝（a-AlOx）的机器学习原子间相互作用势能模型（MLIP）。该数据集由41,203个结构的DFT能量组成，涵盖了不同的超胞大小和K点值。数据集包含了123,560个独特的分子构型，共有4,541,194个原子，包含的元素有铝（Al）和氧（O），属性包括能量、原子力和柯西应力。
The a-AlOx JCP 2020 dataset is used for training the machine learning interatomic potential (MLIP) model for amorphous alumina (a-AlOx). The dataset consists of DFT energies for 41,203 structures, covering various supercell sizes and K-point values. It contains 123,560 unique molecular configurations, with a total of 4,541,194 atoms, including the elements aluminum (Al) and oxygen (O), and properties such as energy, atomic forces, and Cauchy stress.
提供机构：
colabfit
