In addition to removing NSFW images from the training set, this 2.0 release apparently also removed commercial artist styles and celebrities [1]. While it should be possible to fine tune this model to create them anyway using DreamBooth or a similar approach, they clearly went for the safe route after taking some heat.

1. https://twitter.com/emostaque/status/1595731407095140352?s=4...

I predicted back when they started backpedaling that there's a chance that sd1.4 or 1.5 will be the best available model to the general public, for a very long duration, because the backlash will force them to self-castrate themselves.

You can see nobody likes this new model in any of the stable diffusion communities. It's a big flop and for a good reason. The reason it was so successful in the first place was because you could combine artist names to get the model to the outcome you want.

I'll again remind anyone who thinks they might want to use this to download a working version of SD now. They might break their own libraries in the future, and getting SD1.4 could be a real hassle in a year or so. Getting the right .ckpt file, which can have pickled python malware, is not so trivial, and this will get worse in time.

It's going to diverge into castrated official model that intentionally breaks the older models and older models from unofficial shady sources that might contain malware.

As someone completely unfamiliar with SD but interested in playing around with it in the future, what exactly should I download, to have a fully local instance of 1.4 or 1.5?

https://github.com/AUTOMATIC1111/stable-diffusion-webui

https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki...