colabfit/cG-SchNet

cG-SchNet数据集包含了一个在QM9数据集子集上训练的cG-SchNet模型的配置信息。该模型旨在生成具有指定功能团或基序的分子，依赖于分子指纹数据的采样。生成的分子的松弛数据使用ORCA软件计算。数据集包括由cG-SchNet生成的配置的原始数据，以及在不同类型的目标数据上训练的模型和DFT松弛数据作为单独的配置集。数据集大约包含80,000个配置。数据集中的列以dataset_为前缀，存储了额外的详细信息。数据集包含了大约797,272个独特的分子配置，1,467,492个原子，包括C、H、N、O、F元素，并且包含了能量、原子力和Cauchy应力等属性。

The cG-SchNet dataset contains configurations from a cG-SchNet model trained on a subset of the QM9 dataset. The model is intended to generate molecules with specified functional groups or motifs, relying on the sampling of molecular fingerprint data. Relaxation data for the generated molecules is computed using ORCA software. The dataset includes raw data from cG-SchNet-generated configurations, with models trained on several different types of target data and DFT relaxation data as a separate configuration set. It contains approximately 80,000 configurations. Additional details are stored in dataset columns prepended with dataset_. The dataset includes about 797,272 unique molecular configurations, 1,467,492 atoms, comprises elements C, H, N, O, F, and includes properties such as energy, atomic forces, and Cauchy stress.