Too Long; Didn't Read
Run your own text to image prompts with CUDA, a bunch of disk space, and an insane amount of memory. You need a Linux system with a CUDA enabled card to get this working through to the end. There are 2 image generation techniques possible with Latent Diffusion. We will install and take a look at both the txt-2img and txt2-img model that we’ll setup. The RAD model uses a database of images and denises and upscales it to a configurable reference. This seems to mean it’s less good as a general purpose but could be used with a specific training set.