jablonkagroup/MatText-hypo_pot

该数据集包含材料科学领域的多个子数据集，涉及带隙（bandgap）、介电性质（dielectric）、形成能（formenergy）、剪切模量（gvrh）、二维材料（jdft2d）、钙钛矿（perovskites）和声子（phonons）等研究方向。数据集提供了晶体结构的多重表示（如cif文件、z矩阵、原子序列等）以及与能量相关的多种属性（如组成能量、几何能量及其归一化值等）。数据分为训练集和测试集，适用于材料性质预测和机器学习模型的训练与评估。

This dataset comprises multiple sub-datasets in the field of materials science, covering research areas such as bandgap, dielectric properties, formation energy, shear modulus (gvrh), 2D materials (jdft2d), perovskites, and phonons. It provides multiple representations of crystal structures (e.g., cif files, z-matrix, atom sequences) and various energy-related properties (e.g., composition energy, geometry energy, and their normalized values). The data is split into training and test sets, making it suitable for material property prediction and the training and evaluation of machine learning models.