DiscordDiffusion

Discord text-to-image using open source stable-diffusion models.

This uses a smaller model as my VRAM could not handle it. This was also tested targeting cuda:0 which is my first gpu only to let my second gpu run other models using that VRAM.