The Juice is in the description below
Tips to use:
Use LoRA weight between 0.6-0.8
For a basic start use the tags: pikkytingyun, 1girl, solo, looking at viewer, brown hair, red hair, gradient hair, hair between eyes, bangs, long hair, ponytail, green eyes, breasts, large breasts, tail, animal ears, fox ears, fox tail, animal ear fluff, fox girl, detached sleeves, jewelry, bracelet, smile, :d
* The tail is a little hard to make right, it will mostly depent on the base model used *
This LONG prompt are all tags related to her original outfit, If you use theese you will get a pretty accurate result. You can omit some stuff, but i wanted to give all possible tags.
We used one of Meina's model to generate the images.
If you have any questions feel free to ask!
Hope you guys like it! ❤️
Description
FAQ
Comments (3)
A little advice if I may. It appears that you tagged all training images with "fox ears", and therefore generating images without this tag makes the result lack fox ears. It's best to omit tagging features that are immutable parts of the character, this way generated images will have them even when the prompt doesn't mention them
Yea, thats true. But I want my LoRAs to be that way specifically. So the LoRA can both make the character completely but the user also has the choice to omit tags out of it, like the ears or tails so that the featured character doesent have that thing on them. Freedom is key.
Good observation tho!
I disagree - I prefer when things are all tagged, so that I can easily change a character's clothes or features by typing an extra prompt.