A jupyter notebook showing how to finetune the vision transformer on a facial expression dataset (FER-2013)