Abstract: Transfer learning is commonly employed to leverage large, pre-trained models and perform fine-tuning for down-stream tasks. The most prevalent pre-trained models are initially trained using ...