DeePKS-kit is a pure python library so it can be installed following the standard git clone then pip install procedure. Note that the two main requirements pytorch and ABACUS will not be installed automatically so you will need to install them manually in advance. Below is a more detailed instruction that includes installing the required libraries in the environment.
We use conda here as an example. So first you may need to install Anaconda or Miniconda.
To reduce the possibility of library conflicts, we suggest create a new environment (named deepks) with basic dependencies installed (optional):
conda create -n deepks numpy scipy h5py ruamel.yaml paramiko conda activate deepks
Now you are in the new environment called deepks. Next, install PyTorch
# assuming a GPU with cudatoolkit 10.2 support conda install pytorch cudatoolkit=10.2 -c pytorch
Once the environment has been setup properly, using pip to install DeePKS-kit:
$ pip install git+https://github.com/deepmodeling/deepks-kit@abacus
ABACUS with DeePKS enabled
To run DeePKS-kit in connection with ABACUS, users first need to install ABACUS with DeePKS enabled. Details of such installation guide can be found at installation with DeePKS.
While DeePKS-kit has its built-in job dispacther, users are welcome to use DPDispatcher for automatic job submission. The usage of these two types of dispatchers is given in xxx. DPDispacther can simply be installed via
$ pip install dpdispatcher
More details about DPDispacther can be found via DPDispatcher’s documentation.