Installation
ℋ²GB is available for Python>=3.6, and installation is easy!
Note
We do not recommend installation as a root user on your system Python. Please setup a virtual environment, e.g., via Anaconda or Miniconda, or create a Docker image.
Installing ℋ²GB requires external library like PyTorch and PyG. You can easily set up a conda environment to use ℋ²GB. For example, in the following, we create a virtual environment named H2GB:
conda create -n H2GB python=3.9 -y
conda activate H2GB
Then, Install the required packages using the following commands:
conda install pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia
conda install pyg -c pyg
Now you are ready for installing ℋ²GB!
Installation via PyPi
To install ℋ²GB, the easiest way is to simply run:
pip install torch_geometric
Installation from Github
In case a specific version of the dependency is not supported, you can alternatively install by cloning the original Github repository:
git clone https://github.com/junhongmit/H2GB.git