Dataset for UNEP-v1

Dataset for the paper: General-purpose machine-learned potential for 16 elemental metals and their alloys

UNEP-v1 = version 1 of Unified NeuroEvolution Potential as implemented in GPUMD (https://gpumd.org/)

INCAR: Input file of VASP for single-point DFT calculations

nep.in: Input file of GPUMD for NEP training

UNEP-v1-main.txt: The main UNEP-v1 model used in the paper

ensemble-nep-models.zip: The ensemble of eight UNEP-v1 models

trainset.zip: Training data

1-component: all the 1-component training data in extended XYZ format

2-component: all the 2-component training data in extended XYZ format

combined-train-xyz: all the training data in a single extended XYZ file

testset.zip: Test data for Fig 2 and Figs. S1-S17

本数据集配套论文：《面向16种单质金属及其合金的通用机器学习势能》
UNEP-v1：GPUMD（https://gpumd.org/）工具包中实现的统一神经进化势能（Unified NeuroEvolution Potential）第一版
INCAR：用于VASP单点密度泛函理论（Density Functional Theory，DFT）计算的输入文件
nep.in：GPUMD用于神经进化势能（NeuroEvolution Potential，NEP）模型训练的输入文件
UNEP-v1-main.txt：论文中使用的主UNEP-v1模型文件
ensemble-nep-models.zip：包含8个UNEP-v1模型的集成模型压缩包
trainset.zip：训练数据集
1-component：所有以扩展XYZ格式存储的单组分训练数据
2-component：所有以扩展XYZ格式存储的双组分训练数据
combined-train-xyz：整合为单个扩展XYZ文件的全部训练数据
testset.zip：用于生成图2及补充图S1~S17的测试数据集