Show HN: Consumerizing Stable Diffusion

After generating our first image with Stable Diffusion, it was almost impossible to not drop everything we were doing and bust out a simple interface to consumerize the magic of this AI.

Upload your face, train a model, and make great prompts. The challenge we faced was scaling cloud GPUs for each user who trains a model while still keeping a profit margin. But we'd like to think we did it (to some degree), and would love for Hacker News and those that haven't already created a model to give it a go.

Will be a fun way to spend part of your Thanksgiving :)