Another silly LORA that is not that good than I expected... sry..
This lora was trained on ibune_vrc style using public images, and is only for scientific research.
Description
Still learning how to properly train a Lora so this time it's more sensitive and sometimes requires adding "monochrome" into a positive tags (since it was basically trained on black and white images)
Gives best results using "1024x1536" resolution