We use Stable Diffusion v1.5 models, [Open AI's CLIP model ](https://github.com/openai/CLIP), and COCO 2014 caption dataset. Do we need to put their licenses?