Allegro machine learned interatomic potential for silica up to 15000 K

This repository contains the data, scripts, and configurations used for the active learning of a machine learning interatomic potential (using the Allegro architecture) for amorphous silica, trained on DFT data at the r2SCAN level of theory. Overview We train an interatomic potential for silica by iteratively sampling configurations using molecular dynamics and retraining the model on an increasingly large set of structures. The reference electronic structure calculations were performed using VASP with the r2SCAN meta-GGA functional. File Structure 20241108_silica_active_learning_r2scan_v4.ipynb: The main driver script for the active learning loop. It defines the initial systems, the VASP calculator settings, the NequIP/Allegro trainer settings, and the LAMMPS sampler settings. It manages the iterative process of training, sampling, and labeling. compute_single_atom_energies.ipynb: A standalone script used to compute the reference energies of isolated Silicon and Oxygen atoms. These energies are necessary for normalizing the total energies during model training. allegro_input_v4.yaml: The configuration file for the Allegro model architecture and training hyperparameters. active_learning_analysis.ipynb: A notebook for analyzing the progress of the active learning loop, including training/validation set growth, parity plots (energy, forces, virial), and uncertainty calibration analysis. sampler_analysis.ipynb: A notebook for analyzing the sampling phase, specifically checking LAMMPS log files for simulation temperatures, lengths, and halt conditions. al_out_allegro_v4/: The main output directory containing the data and results for each iteration of the active learning loop. This includes: Trained models Sampled structures VASP calculation outputs Training/validation datasets encut_sensitivity_analysis/: Contains VASP calculations and an analysis notebook (dataset_analysis.ipynb) performing convergence tests for the plane-wave energy cutoff (ENCUT). Calculations are performed at varying ENCUT values (e.g., 800, 1000, 1200, 1400, 1600 eV) to assess the impact on pressure (Pulay stress) and forces, justifying the choice of 1000 eV for the main dataset generation. initial_system_generation/: Contains the initial structures used to start the active learning process (e.g., alpha-quartz, beta-cristobalite, coesite, and amorphous structures). single_atom_energies/: Output directory for the single atom energy calculations. templates/: LAMMPS input templates used for various sampling strategies (melt, quench, shear, tension). silica_fracture_data/: data + scripts to reproduce crack velocity, fracture surface energy, crack tip temperature and eleastic properties figures of the paper about fracture simulations using the present silica potential Methodology Active Learning Workflow The workflow is orchestrated by the 20241108_silica_active_learning_r2scan_v4.ipynb notebook using the hyal framework. The methodology is based on the approach described in: Cezar, H. M., Bodenstein, T., Sveinsson, H. A., et al. "Learning atomic forces from uncertainty-calibrated adversarial attacks." npj Comput Mater 11, 200 (2025). https://doi.org/10.1038/s41524-025-01703-5 Key steps include: Initialization: Starting with a set of crystalline silica structures. Training: An ensemble of Allegro models is trained on the available labeled data. Sampling: The ensemble is used to drive Molecular Dynamics (MD) simulations (using LAMMPS). The simulations employ various protocols (melt, melt-quench, mechanical deformation) to explore the configuration space. Configurational refinement: Selected configurations are driven to moderate model disagreement (uncertainty) . Labeling: The selected configurations are calculated using VASP (r2SCAN) to obtain ground truth energies, forces, and virials. Loop: The new data is added to the training set, and the cycle repeats. Electronic Structure Details (VASP) Functional: r2SCAN (Meta-GGA) Pre-convergence: PBE (relaxed constraints) Plane Wave Cutoff: 1000 eV Precision: Accurate Machine Learning Model (Allegro) Architecture: Allegro (E(3)-equivariant local interaction neural network) Cutoff Radius: 5.5 Å L_max: 1 (Full O(3) parity) Layers: 1 tensor product layer Features: 64 tensor features Optimization: Adam optimizer with ReduceLROnPlateau scheduler Requirements The scripts rely on several Python packages and external software: Python Libraries: hyal, hyif, hyset (internal/project-specific wrappers), ase, nequip, numpy, pandas, matplotlib, lammps_logfile. External Software: VASP (for DFT calculations) LAMMPS (with Allegro/NequIP pair style support for MD sampling) Single Atom Energies The compute_single_atom_energies.ipynb script calculates the reference energies for isolated atoms. Box size: 15.0 x 15.0 x 15.0 Å Species: Si, O Method: Same VASP settings as the main active learning loop.本仓库包含用于无定形二氧化硅机器学习原子间势（采用Allegro架构）的数据集、脚本与配置文件，该势函数基于r2SCAN理论水平下的密度泛函理论（Density Functional Theory, DFT）数据训练得到。 研究概述：我们通过分子动力学迭代采样构型，并在逐步扩大的结构数据集上重新训练模型，以此构建二氧化硅原子间势函数。参考电子结构计算采用VASP软件结合r2SCAN meta-广义梯度近似（meta-Generalized Gradient Approximation, meta-GGA）泛函完成。 文件结构 20241108_silica_active_learning_r2scan_v4.ipynb：主动学习循环的核心驱动脚本，用于定义初始体系、VASP计算参数设置、NequIP/Allegro训练超参数以及LAMMPS采样器配置，并管理训练、采样与标注的迭代流程。 compute_single_atom_energies.ipynb：用于计算孤立硅、氧原子参考能量的独立脚本，该能量可用于模型训练过程中总能量的归一化处理。 allegro_input_v4.yaml：Allegro模型架构与训练超参数的配置文件。 active_learning_analysis.ipynb：用于分析主动学习循环进展的Jupyter Notebook，涵盖训练/验证集规模增长、奇偶图（能量、力与应力张量）以及不确定性校准分析等内容。 sampler_analysis.ipynb：用于分析采样阶段的Notebook，重点针对LAMMPS日志文件检查模拟温度、时长与终止条件。 al_out_allegro_v4/：主动学习各迭代阶段的数据与结果主输出目录，包含以下内容： - 训练完成的模型 - 采样得到的构型 - VASP计算输出文件 - 训练/验证数据集 encut_sensitivity_analysis/：包含VASP计算与分析脚本dataset_analysis.ipynb，用于开展平面波能量截断能（ENCUT）的收敛性测试。测试采用800、1000、1200、1400、1600 eV等多种截断能取值，评估其对压强（Pulay应力）与原子力的影响，以此论证主数据集生成阶段选用1000 eV截断能的合理性。 initial_system_generation/：包含启动主动学习流程所需的初始构型，例如α-石英、β-方石英、柯石英以及无定形结构。 single_atom_energies/：孤立原子能量计算的输出目录。 templates/：适用于多种采样策略（熔融、淬冷、剪切、拉伸）的LAMMPS输入模板文件。 silica_fracture_data/：包含复现论文中关于采用本二氧化硅势函数开展断裂模拟的相关数据与脚本，用于生成裂纹扩展速率、断裂表面能、裂纹尖端温度与弹性性能相关图表。 研究方法 主动学习工作流程 本工作流由20241108_silica_active_learning_r2scan_v4.ipynb基于hyal框架实现，其方法论参考如下文献： Cezar, H. M., Bodenstein, T., Sveinsson, H. A., et al. "Learning atomic forces from uncertainty-calibrated adversarial attacks." npj Comput Mater 11, 200 (2025). https://doi.org/10.1038/s41524-025-01703-5 核心步骤包括： 1. 初始化：以一组结晶态二氧化硅结构作为初始数据集。 2. 训练：基于现有标注数据训练Allegro模型集成。 3. 采样：利用模型集驱动分子动力学（Molecular Dynamics, MD）模拟（采用LAMMPS软件），模拟过程涵盖熔融、熔融淬冷、机械形变等多种协议以探索构型空间。 4. 构型优化：将选中的构型调整至中等模型分歧（不确定性）水平。 5. 标注：采用VASP（r2SCAN泛函）对选中构型进行计算，获取真实的能量、原子力与应力张量数据。 6. 循环：将新标注数据加入训练集，重复上述迭代流程。 电子结构计算细节（VASP） - 泛函：r2SCAN（meta-广义梯度近似泛函） - 预收敛：采用PBE泛函进行约束弛豫 - 平面波截断能：1000 eV - 计算精度：高精度模式 机器学习模型（Allegro） - 架构：Allegro（E(3)等变局域相互作用神经网络） - 截断半径：5.5 Å - L_max：1（完整O(3)宇称对称性） - 层数：1个张量积层 - 特征维度：64个张量特征 - 优化器：Adam优化器，搭配ReduceLROnPlateau学习率调度器 依赖项 本脚本依赖多个Python库与外部软件： Python库：hyal、hyif、hyset（内部项目专属封装库）、原子模拟环境（Atomic Simulation Environment, ASE）、nequip、numpy、pandas、matplotlib、lammps_logfile。 外部软件： - VASP：用于DFT计算 - LAMMPS：需支持Allegro/NequIP对势风格，用于分子动力学采样 孤立原子能量计算 compute_single_atom_energies.ipynb脚本用于计算孤立原子的参考能量： - 模拟盒子尺寸：15.0 × 15.0 × 15.0 Å - 原子种类：Si、O - 计算参数：与主主动学习循环保持一致的VASP设置。