Model can be run in inference mode just by rinning main.py as it is. There is a setting section at the top of the file where the model can be toggled to retrain, save etc.