一、module命令module purge module load XXX module unload XXX二、机群上的模块华东四区山东#华东四区山东------------------------- /work/home/ln2026/apprepo/modules -------------------------- lammps-22Jul2025-oneapi2024.0.1 --------------------------- /public/software/modules/apps ---------------------------- abinit/8.10.3-hpcx-intel-2017 alphafold2/dtk22.04-DCU amber/2018-hpcx-gcc-7.3.1 AutoDock/4.2.6-DCU2-hpcx-gcc-7.3.1 bigDFT/1.8.3-hpcx-intel-2017 Calculix/2.16-hpcx-gcc-7.3.1 castep/22.11/ifort-mpi/linux_x86_64_ifort--mpi cdo/1.7.2-intel-2017 CMAQ/5.3-hpcx-intel-2017 Code_Aster/13.8-hpcx-gcc-7.3.1 COMSA/2.5.1-DCU2-hpcx-gcc-7.3.1 cp2k/2023.1-intelmpi-2018 cp2k/2023.1-intelmpi-2018-plumed cp2k/2023.1-intelmpi-2021 cp2k/2023.2-gcc-930 cp2k/2023.2-intelmpi-2021-plumed cp2k/2024.1-gcc-930 cp2k/4.1-intelmpi-2017 cp2k/5.1-intelmpi-2017 cp2k/6.1.0-hpcx-intel-2017 cp2k/6.1.0-intelmpi-2017-plumed cp2k/7.1-intelmpi-2018 cp2k/7.1-intelmpi-2020 cp2k/8.1-intelmpi-2020 cp2k/8.2-intelmpi-2020 cp2k/9.1-gcc-930 cp2k/9.1-intelmpi-2018 cp2k/9.1-intelmpi-2020 cp2k/9.1-intelmpi-2021 curl/7.58.0-intel-2017 dl-poly/5.1.0-intelmpi-intel-2021 DUNE/2.8-hpcx-gcc-7.3.1 Elmer/8.4-hpcx-intel-2017 FlameGraph/master fvcom/3.2-hpcx-intel-2017 gamess-us/2019R2-hpcx-intel-2017 Genomeworks/0.4.0-DCU2-hpcx-gcc-7.3.1 git/2.30.2-gcc-7.3.1 g_mmpbsa/1.6-gcc-7.3.1 gnuplot/5.0.5-gcc-7.3.1 grads/2.2.0-prebuild grads/2.2.1-gcc-4.8.5 gromacs/2016.6-intelmpi-2021 gromacs/2018.8-intelmpi-2021 gromacs/2019.5-hpcx-intel-2017 gromacs/2019.6-intelmpi-2021 gromacs/2020.1-DCU2-hpcx-gcc-7.3.1 gromacs/2020.3-DCU2-hpcx-gcc-7.3.1 gromacs/2020.6-intelmpi-2021 gromacs/2021.3-intelmpi-2021 gromacs/2021.6-intelmpi-2021 gromacs/2022.1-DCU2-gcc-7.3.1 gromacs/2022.1-DCU2-hpcx-gcc-7.3.1 gromacs/2022.1-DCU2-hpcx-gcc-7.3.1-bak gromacs/2022.2-hpcx-gcc-7.3.1 gromacs/2023.1-gcc930-intelmpi21 gromacs/4.5.6-intelmpi-2017 gromacs/4.6.7-intelmpi-2021 gromacs/5.0.7-intelmpi-2021 gromacs/5.1.5-intelmpi-2021 gromacs-plumed/2019.6-intelmpi-2021 gromacs-plumed/2020.6-intelmpi-2021 gromacs-plumed/2021.4-intelmpi-2021 Hashcat/6.1.1-DCU2-hpcx-gcc-7.3.1 Hoomd-blue/3.0.0-DCU2-hpcx-gcc-7.3.1 ImageMagick/7.0.10-prebuild IPM/2.0.6-hpcx-gcc-7.3.1 IPM/2.0.6-impi-2017 jdk/19.0.2 lammps/16Mar18-intelmpi-2017 lammps/22Aug18-intelmpi-2017 lammps/23Jun2022-DCU2-hpcx-gcc-7.3.1 lammps/23Jun2022-intel_21 lammps/24Dec2020-DCU2-hpcx-gcc-7.3.1 lammps/24Dec20-intelmpi-2017 lammps/28Mar2023-intel_21 lammps/29Oct20_intel17 lammps/30Jul21-intelmpi-2017 lammps/30Jul21-intelmpi-2021 lammps/3Mar20-intelmpi-2017 lammps/5Jun19-intelmpi-2017 lammps/7Aug19-hpcx-intel-2017 lammps/7Aug19-intelmpi-2017 lammps/8Apr21-intelmpi-2017 MiniFE/2.1.0-hpcx-intel-2017 MISA-AKMC/v0.1.0-beta-hpcx-gcc-7.3.1 MISA-MD/v0.4.0-hpcx-gcc-7.3.1 MISA-SCD/v1.0.0-hpcx-gcc-7.3.1 NAMD/2.13-hpcx-intel-2017 NAMD/2.15-DCU2-gcc-7.3.1 NAMD/2.15-DCU2-hpcx-gcc-7.3.1 ncl_ncarg/6.5.0-gcc-4.8.5 nco/4.6.7-intel-2017 nco/4.6.7-intel-2021 ncview/2.1.7-intel-2017 nwchem/6.8.1-hpcx-intel-2017 octave/5.2.0-gcc-7.3.1 octave/5.2.0-gcc-7.3.1-mt octopus/8.4-hpcx-intel-2017 OpenFOAM/10-openmpi415-gcc-9.3.0 OpenFOAM/6-hpcx-gcc-7.3.1 OpenFOAM/7-hpcx-gcc-7.3.1 OpenFOAM/8-openmpi415-gcc-9.3.0 OpenFOAM/9-openmpi415-gcc-9.3.0 OpenFOAM/v1812-hpcx-gcc-7.3.1 OpenFOAM/v1912-hpcx-gcc-7.3.1 OpenFOAM/v2012-hpcx-gcc-7.3.1 OpenFOAM/v2112-hpcx-gcc-7.3.1 OpenFOAM/v2212-hpcx-gcc-7.3.1 opengrads/2.0.a8-intel-2017 OpenMM/7.5.0-DCU2-hpcx-gcc-7.3.1 OpenMM/8.0.0-DCU2-hpcx-gcc-7.3.1 openmx/3.9.2-hpcx-intel-2017 ovito/3.8.4 p7zip/9.20.1 ParaView/5.8.0-prebuild PhengLei/v2306/intel2017 PhengLei/v2306/mpich3.3.2-gcc-7.3.1 ploticus/2.42-prebuild plumed/2.8.0-intel-2021 proj/4.9.2-intel-2017 pwkit/2.0 pwmat/20240223-DCU2-intelmpi-2021 pwmat/20240223-magma-DCU2-intelmpi-2021 python/3.8.10 Pytorch/1.10.0-hpcx-gcc-7.3.1 Pytorch/1.8.0-hpcx-gcc-7.3.1 qe/5.4.0-intelmpi-2017 qe/6.1-intelmpi-2017 qe/6.4.1-hpcx-intel-2017 qe/6.4.1-intelmpi-2017 qe/6.4-intelmpi-2017 qe/6.5-hpcx-intel-2017 qe/6.5-intelmpi-2017 qe/6.6-hpcx-intel-2017 qe/6.6-intelmpi-2017 qe/6.7-intelmpi-2017 qe/6.8-hpcx-intel-2017 qe/6.8-intelmpi-2017 qmcpack/3.9.1-hpcx-gcc-7.3.1 ROMS/3.6-hpcx-intel-2017 siesta/4.0.2-hpcx-intel-2017 SU2/7.0.1-hpcx-gcc-7.3.1 Tensorflow/1.15.1-cp36 tmux/1.8 tmux/3.2 unidock/dtk23.10-DCU vaspkit/1.2.4-prebuild vtstscripts/1033 WPS/3.9.1-hpcx-intel-2017 WRF/3.8-hpcx-intel-2017 WRF/3.9.1-hpcx-intel-2017 WRF/4.2-hpcx-intel-2017 --------------------------- /public/software/modules/base ---------------------------- anaconda2/4.3.0 mathlib/jasper/1.900.1-intel-2017 anaconda3/2023.09 mathlib/jasper/1.900.1-intel-2021 anaconda3/5.2.0 mathlib/lapack/3.8.0-gcc-7.3.1 apps/apptainer/1.2.4 mathlib/lapack/3.8.0-intel-2017 apps/apptainer/1.3.4 mathlib/libjpeg/9.1.0-intel-2017 apptainer/1.2.4 mathlib/libjpeg/9.1.0-intel-2021 automake/1.16.1 mathlib/libpng/1.2.50-gcc-7.3.1 compiler/cmake/3.15.6 mathlib/libpng/1.2.50-intel-2017 compiler/cmake/3.20.4 mathlib/libpng/1.2.50-intel-2021 compiler/cmake/3.23.3 mathlib/libxc/4.3.4-intel-2017 compiler/devtoolset/7.3.1 mathlib/netcdf/3.6.3-intel-2017 compiler/dtk/1 mathlib/netcdf/4.4.1-intel-2017 compiler/dtk/21.04 mathlib/netcdf/4.6.1-intel-2021 compiler/dtk/22.04.2 mathlib/netcdf/c4.9.2_f4.6.1-gcc731 compiler/dtk/22.10 mathlib/netcdf/c4.9.2_f4.6.1-intel21 compiler/dtk/22.10.1 mathlib/opencv/3.1.0-gcc-7.3.1 compiler/dtk/23.04 mathlib/openjpeg/2.3.1-intel-2017 compiler/dtk/23.04.1 mathlib/petsc/3.8.3-gcc7.3.1_hpcx compiler/dtk/23.10 mathlib/pnetcdf/1.12.0-gcc-7.3.1 compiler/dtk/24.04 mathlib/pnetcdf/1.12.0-impi-2017 compiler/dtk/26.04 mathlib/pnetcdf/1.12.1-hpcx-intel-2017 compiler/gcc/11.4.0 mathlib/pnetcdf/1.12.2-impi-2021 compiler/gcc/12.2.0 mathlib/pnetcdf/1.8.1-hpcx-intel-2017 compiler/gcc/13.3.0 mathlib/scalapack/2.0.0-intel-2017 compiler/gcc/9.3.0 mathlib/szip/2.1.1-intel-2017 compiler/go/1.16.4 mathlib/udunits/2.2.25-intel-2017 compiler/intel/2013.2.146 mathlib/udunits/2.2.26-intel-2017 compiler/intel/2015.2.164 mathlib/zlib/1.2.11-intel-2017 compiler/intel/2017.5.239 mathlib/zlib/1.2.8-gcc-7.3.1 compiler/intel/2018.5.274 mathlib/zlib/1.2.8-intel-2017 compiler/intel/2020.1.217 mathlib/zlib/1.3-intel-2021 compiler/intel/2021.3.0 mpi/hpcx/2.4.1-gcc-7.3.1 compiler/pgi/20.9 mpi/hpcx/2.4.1-intel-2017 compiler/rocm/3.3 mpi/hpcx/2.7.4/intel-2017.5.239 mathlib/antlr/2.7.7-intel-2017 mpi/hpcx/2.7.4-gcc-7.3.1 mathlib/blas/3.5.0-intel-2017 mpi/hpcx/gcc-7.3.1 mathlib/boost/1.67.0-intel-2017 mpi/hpcx/intel-2017.5.239 mathlib/elpa/2019.05.001-impi-2020 mpi/intelmpi/2017.4.239 mathlib/fftw/3.3.10_intel21_double mpi/intelmpi/2018.4.274 mathlib/fftw/3.3.10_intel21_single mpi/intelmpi/2020.1.217 mathlib/fftw/3.3.10-intelmpi_21-double mpi/intelmpi/2021.3.0 mathlib/fftw/3.3.10-intelmpi_21-single mpi/intelmpi/5.0.2.044 mathlib/fftw/3.3.8-gcc-7.3.1-double mpi/mpich/3.1.3-intel2013 mathlib/fftw/3.3.8-gcc-7.3.1-single mpi/mpich/3.3.2-gcc-7.3.1 mathlib/fftw/3.3.8-intel-2017-double mpi/mpich/4.1.2-gcc-9.3.0 mathlib/fftw/3.3.8-intel-2017-single mpi/openmpi/3.1.4-intel-2017 mathlib/grib2/2.0.7-intel-2017 mpi/openmpi/openmpi-3.1.6 mathlib/grib_api/1.28.0-intel-2017 mpi/openmpi/openmpi-4.1.5-gcc9.3.0 mathlib/gsl/2.6-intel-2017 nvidia/cuda/11.3 mathlib/hdf4/4.2r1-intel-2017 nvidia/cuda/11.6 mathlib/hdf5/1.12.0-intel-2021 nvidia/cuda/11.7 mathlib/hdf5/1.12.2-gnu-7.3.1 nvidia/cuda/11.8 mathlib/hdf5/1.12.2-gnu-7.3.1-parallel nvidia/cuda/12.1 mathlib/hdf5/1.12.2-intel-21 nvidia/cuda/12.2 mathlib/hdf5/1.8.20-intel-2017 nvidia/nvhpc/20.9 mathlib/htslib/1.9-gcc-7.3.1 R/3.6.3-gcc-7.3.1 mathlib/ioapi/3.2-intel-2017 singularity/3.7.3 mathlib/jasper/1.900.1-gcc-7.3.1三、安装软件提醒先加载依赖项然后推荐使用conda安装而非pip install。conda install 与 pip install 的区别在conda 环境里用 pip install 只是把包塞进这个环境的site-packages而 conda install 不仅装包还会在整个conda 生态里做严格的依赖解析、版本锁定、二进制兼容检查并在需要时降级/升级你环境中的其它conda 包。一句话总结conda 是跨语言包管理器管包 环境 底层 C/C 依赖pip 是 Python 专属包管理器只管 Python 库有conda install 就先用没有的话 再用 pip install四、以NequIP为例安装先看手册需要python和Pytorchconda create-n nequip python3.11-y conda activate nequip module load gcc $ git clone--depth1https://github.com/mir-group/nequip.git $ tar-zxvf nequip-0.19.0.tar.gz $ cd nequip-0.19.0/$ pip install.pip install . 在这里做了这几件事读取构建配置找到目录里的 pyproject.tomlNequIP 用它来声明依赖、构建方式等。解析依赖根据配置文件检查并安装 NequIP 所需的依赖如 torch、e3nn、ase、hydra-core 等。构建项目如果项目包含需要编译的 C/C/CUDA 扩展NequIP 有少量性能相关的 C 扩展会用你当前环境的编译器你之前设置的 GCC 11.4.0进行编译。安装到环境把构建好的包复制到当前 Python 环境的 site-packages 目录即你之前看到的 /work/home/ln2026/miniconda3/envs/nequip/lib/python3.11/site-packages/并生成可执行脚本如 nequip-train、nequip-evaluate 等命令行工具。总结经验缺少补啥配置环境变量配好之后测试一下通过算例。写个脚本。因为网络问题推荐使用源码安装。