I made these for quick architectural visualization either by
They might or might not work well with normal txt2img generation.
The models are supposed to know about more architectural details than base checkpoints.
I have some more in the pipeline. Will upload if there's interest.