On the Language Encoder of Contrastive Cross-modal Models

Open in new window