CLIP by itself is simply the primary practical iteration of translating text-to-photographs, and I suspect this gained’t be the final implementation of this type of model (OpenAI could pull a GPT-3 rather than open up-source the inevitable CLIP-2 considering that now there’s a confirmed monetizeable use situation). The notebook http://aiartprompt24578.is-blog.com/15170608/getting-my-ai-art-prompt-to-work