This lora is intended to make mimics out of any object! (results may vary)
Trigger Word: m1m1c aesthetic
This lora is not strictly needed, but it's a big help for Moss. The model imparts a bias towards an illustrated concept art style, but it's a weak bias and easy to overcome with prompts or loras. You can customize the mimics pretty easily. I've been using variations of 'Monstrous grinning mouth with sharp teeth and a long tongue. It has many randomly placed eyes'.
Note: I trained using trigger word m1m1c. If you use that alone, it will literally put m1m1c text in the image at random places. To combat that I associate the trigger word with a directive... in this case a fairly meaningless one 'm1m1c aesthetic'. I found 'm1m1c style' also works to the correct the issue, and I'm sure there's many others combinations.



















