[Future Proofing] Preparing for v1.5 models

[JoshuaKimsey]

I'm seeing an increasing number of posts on the current open beta for the v1.5 Stable Diffusion models. They look really good, as you can see from this and this Reddit posts. v1.5 looks really impressive with what it's achieving, and while people are speculating it'll be days or weeks away, it definitely seems to be on its way.

With that, I thought I would go ahead and suggest that a system be made to allow people to switch between models when running Stable Diffusion. Either dynamically in the webui, if that is possible, or at launch time. I've already got the early basis of support for loading new models in my script, even though it is disabled currently. However, if multiple models becomes a legitimate thing people wish to switch between, we may need a way to easily sort between models, as currently it's just loaded as model.ckpt.

It's still likely days, or even weeks, out from becoming something that would be needed, but figured it was probably a good idea to go ahead and suggest this so it will be ready come the v1.5 model launch.