Here's my second Lora. I know there are already a few loras for dirty feet and socks, but somehow they don't fit my needs for creating AI-Stuff with it. So I made one myself.

Trained on more than 600 handpicked pictures, which are mostly high quality. Took me about five hours to gather them and put prompts on them.

Start by using a strength of 0.6.

Going to 1.0 or above will generate very realisitc pictures, so keep it lower, if you don't want to ruin the style, you're looking for.

This Lora ist trained on realistic pictues, so use an realistic model, or Toonbabes. Toonbabes works superb as well. Haven't tested more, for now.

Feel free to send me feedback, so I can improve it.

Trained on 16 epochs and 9,712 steps.