Installation

1. Install with conda environment configuration file

• Download the agat_linux_gpu_cu124.yml or agat_win_cpu.yml file.

• Run

  conda env create -f agat_linux_gpu_cu124.yml
  

• Run conda env list to check installed environments.

2. Install manually with conda

Examined dependency compatibility:

OS	Python	DGL	PyTorch	numpy	ASE	CUDA
Windows CPU	3.10	2.2.1	2.3.0	2.0.1	3.23.0	-
Linux GPU	3.12	2.4.0	2.4	2.0.1	3.23.0	12.4

3. Install with Conda environment on Linux.

• Create a new environment

  conda create -n agat python==3.12
  

• Activate the environment

  conda activate agat
  

• Install PyTorch, Navigate to the installation page and choose your platform. For example (GPU):

  pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu126
  

• Install dgl. Please navigate to the Get Started page of dgl. For example (GPU):

  pip install  dgl -f https://data.dgl.ai/wheels/cu124/repo.html
  

• Install AGAT package

  pip install agat
  

• Install CUDA and CUDNN [Optional].

  • For HPC with Linux OS, you may load CUDA by checking module av, or you can contact your administrator for help.

  • Or download manually:

    • CUDA Toolkit

    • cuDNN

4. Install with Conda environment on Windows.

The DGL package has limited compatibility with Windows, particularly for CUDA versions. You are highly recommended to use the CPU version on Windows.

• Create a new environment

  conda create -n agat python==3.10
  

• Activate the environment

  conda activate agat
  

• Install PyTorch,
  Navigate to the installation page and choose your platform. For example (GPU):

  conda install pytorch==2.3.0 torchvision==0.18.0 torchaudio==2.3.0 cpuonly -c pytorch
  

• Install packaging: pip install packaging

• Install dgl.

• Install ASE: pip install ase.

• Install AGAT package

  pip install agat