One should follow same procedure as in the original framework provided by https://github.com/TRI-ML/permatrack/tree/main. Instead of the Docker file we provide a conda environment that runs on the lisa-cluster environment, as the original Docker file was not cluster-friendly in our case. After installing the conda environment (gpu_env.yml) proceed to step 2 of the INSTALL.md of permatrack and continue as explained. We would like to give a special token of appreciation to the authors of permatrack for their contribution to this field.