This is my 2nd LoRA and it is trained on a small set of RimWorld screenshots (without game GUI).

The results are a little disappointing, but the straight lines and sharp shapes are far from what SDXL usually generates I assume. It tends to (wrongly) select an isometric perspective every 3 tries or so.

I still really enjoy the style of the output, so decided to share it anyway. Hope someone will enjoy it for inspiration perhaps!

Using 'RimWorld' in your prompt helps a lot.