Supervised Fine-tuning in turn Improves Visual Foundation Models

Open in new window