This was a Lora I made quite a while ago as an experiment.

I only used five photos that I had taken myself, and then trained a Lora based on what AI said all my photos had in common.

I suppose this could be considered a liminal space lora.

Look at the prompts used in my images for example language used.

It might even work better if you ignore the way I captioned it so you can get more diverse output.