This is my Mann-E platform, it is an attempt to create a large ai text-to-image model and it's based on Stable Diffusion. But you know what? I think I had "self instruct" way before Stanford, 'cause the input data was from Dall-E and Midjourney!