This model needs a trigger in your prompt: pipedisplay

It is difficult to train and seems to be slightly over done.

Hopefully you can enjoy it for now until I am able to find a better training of this.

The images aren't cherry picked - these are just four of the first ones that I generated.

I wanted to get this out because it seems fun! Haven't really tested but I suspect that it'll work best at low guidance (less than 3.5).

If you get interesting results with this please share!