TOV-VICReg Pretraining the Vision Transformer using self-supervised methods for vision based Deep Reinforcement Learning pip install -e . cd tov_vicreg/models ./run_tov-vicreg.sh