Image embedding generation service based on the publication CoCa: Contrastive Captioners are Image-Text Foundation Models. Very strong image-to-text model, capable of generating captions for images. Further investigation on the model will be handled in upcoming projects.