I started out by using Easy Diffusion. It uses microsoft's DirectML and is really slow but it was very easy to use and learn.
Now I am using ComfyUI. It's using ROG instead which is orders of magnitude faster (now I can do images higher than 768x768). It's also very configurable.
