AI4S/REANN
模型介绍文件和版本Pull Requests讨论分析
下载使用量0

递归嵌入原子神经网络

组件版本

hdk: 25.0.rc1.1
cann：8.3.rc1
torch：2.1.0
torch_npu：2.1.0.post17

环境准备

  • 创建容器：根据昇腾机器版本切换基础镜像，镜像列表链接：https://www.hiascend.com/developer/ascendhub/detail/17da20d1c2b6493cb38765adeba85884
docker run -it  -u root \
 --net=host --shm-size=5g \
 --device=/dev/davinci_manager \
 --device=/dev/hisi_hdc \
 --device=/dev/davinci4 \
 --device=/dev/davinci5 \
 -v /usr/local/dcmi:/usr/local/dcmi \
 -v /usr/local/bin/npu-smi:/usr/local/bin/npu-smi \
 -v /usr/local/Ascend/driver/lib64/:/usr/local/Ascend/driver/lib64/ \
 -v /usr/local/Ascend/driver/version.info:/usr/local/Ascend/driver/version.info \
 -v /etc/ascend_install.info:/etc/ascend_install.info \
 -v /usr/share/zoneinfo/Asia/Shanghai:/etc/localtime \
 -v /home:/home/ \
 --name REANN_test \
 --entrypoint=/bin/bash \
 -it swr.cn-south-1.myhuaweicloud.com/ascendhub/cann:8.3.rc1-a3-ubuntu22.04-py3.11 # 根据昇腾机器版本切换基础镜像
  • 安装相关依赖

安装torch和torch_npu

wget https://gitcode.com/Ascend/pytorch/releases/download/v7.2.0-pytorch2.1.0/torch_npu-2.1.0.post17-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
pip3 install torch_npu-2.1.0.post17-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
pip3 install protobuf==3.20
  • 安装三方依赖库
pip3 install numpy=1.24.0
pip3 install opt-einsum==3.2.0
  • 下载代码
git clone https://atomgit.com/AI4Science/REANN.git
cd REANN/reann
  • 测试数据准备
mkdir para && cd para
ln -s ../../example/co2+ni100/para/input_density ./input_density
ln -s ../../example/co2+ni100/para/input_nn ./input_nn
cd ../../data
cp -r co2+Ni100/ co2+ni100/
  • 训练
python3 -m torch.distributed.run --master_addr 127.0.0.1 --master_port 12345 --nproc_per_node=1 --nnodes=1  --standalone ./
  • 其他问题 可能运行集群化训练遇到如下问题，可以在/etc/hosts中添加ipv6域名映射：127.0.0.1 hostname-qdcab.foreman.pxe alt text

参考文献

如果您使用此软件包，请引用以下文献。

  1. 原始 EANN 模型：Yaolong Zhang, Ce Hu and Bin Jiang J. Phys. Chem. Lett. 10, 4962-4967 (2019).
  2. 用于偶极矩/跃迁偶极矩/极化率的 EANN 模型：Yaolong Zhang, Sheng Ye, Jinxiao Zhang, Jun Jiang and Bin Jiang J. Phys. Chem. B 124, 7284–7290 (2020).
  3. REANN 模型理论：Yaolong Zhang, Junfan Xia and Bin Jiang Phys. Rev. Lett. 127, 156002 (2021).
  4. REANN 实现细节：Yaolong Zhang, Junfan Xia and Bin Jiang J. Chem. Phys. 156, 114801 (2022).