CoNbV CMS2019

This dataset was generated using the following active learning scheme: 1) candidate structures were relaxed by a partially-trained MTP model, 2) structures for which the MTP had to perform extrapolation were passed to DFT to be re-computed, 3) the MTP was retrained, including the structures that were re-computed with DFT, 4) steps 1-3 were repeated until the MTP no longer extrapolated on any of the original candidate structures. The original candidate structures for this dataset included about 27,000 configurations that were bcc-like and close-packed (fcc, hcp, etc.) with 8 or fewer atoms in the unit cell and different concentrations of Co, Nb, and V.

本数据集采用以下主动学习方案生成：1）候选结构通过部分训练的MTP模型进行松弛处理，2）对于MTP需要执行外推的结构，将其传递至DFT进行重新计算，3）对MTP进行重新训练，包括使用DFT重新计算的结构，4）重复步骤1至3，直至MTP不再对外推任何原始候选结构。本数据集的原始候选结构包括约27,000种配置，这些配置类似于体心立方（bcc）和密堆积（fcc、hcp等），单元细胞中原子数量不超过8个，且Co、Nb和V的浓度各不相同。