Version: 1.0

Installation

On Mcloud

首先加载编译器(以 mcloud 为例)

module load cuda/11.6 intel/2020
source /opt/rh/devtoolset-8/enable

然后利用 conda 创建一个新 python 环境

conda create -n PWMLFF python=3.8

After PWMLFF has been created, re-enter the current environment

conda deactivate
conda activate PWMLFF

Install the following packages

pip install pymatgen scikit-learn-intelex numba tensorboard
# pip uninstall op    # if op has been installed

conda install pytorch==1.12.1 cudatoolkit=11.6 -c pytorch -c conda-forge

Also, make sure your g++ supports C++ 14 standard. Use "g++ -v" to check, and version above 7.0 should be fine. Intel compiler is also required.

然后进入含 GPU 的节点，并申请到 GPU 资源，可通过如下命令确认是否申请到了 GPU

python
>> import torch
>> torch.cuda.is_available()

返回值为 True 则可开始编译。在github 网站下载源码后，进入 ./src 文件夹并开始编译

sh build.sh

If the building is successful, modify the following environment variables

vim ~/.bashrc
export PATH=absolute/path/to/PWMLFF/src/bin:$PATH
export PYTHONPATH=absolute/path/to/PWMLFF/src/:$PYTHONPATH
source ~/.bashrc

Lammps

Linear Model, KFNN, and KFDP¶

MLFF provides an interface for LAMMPS. You should compile LAMMPS from the source code in order to use it. Intel Fortran and C++ compilers are required.

First, obtain LAMMPS's source code and unzip it. https://github.com/LonxunQuantum/Lammps_for_PWMLFF

GNN

If you wish to use GNN for in LAMMPS, see the page below for guidance.

https://github.com/mir-group/pair_nequip

Installation

On Mcloud​

Lammps​

Linear Model, KFNN, and KFDP¶​

GNN​

On Mcloud

Lammps

Linear Model, KFNN, and KFDP¶

GNN