To ease development, we provide a general conda environment with some of the packages that we will use in these tutorials, then we let the user find the PyTorch version that fits their machine requirements, and finally we install the custom packages.
conda create --file environment.yml
Besides, we recomend users to install the PyTorch version matching their system requirements. Check the PyTorch webpage to find how to install PyTorch. We tested that the PyTorch LTS with 11.1 CUDA drivers runs OK.
The tutorial packages can be installed from the source code. We recommend to follow this sequence to avoid some dependencies clashes.
pip install git+https://github.com/brunocuevas/a3md-utils.git
pip install git+https://github.com/brunocuevas/a3md.git
pip install git+https://github.com/brunocuevas/DeepDFT.git
pip install pyscf == 1.7.6
To start the tutorials, we can just open a jupyter-lab session and open the notebooks.