Aligning Visual Contrastive learning models via Preference Optimization

Open in new window