COSMOS: Cross-Modality Self-Distillation for Vision Language Pre-training

Open in new window