EDIT: The Pony model is just a lame attempt at a style lora. use it with other pony models to get the original style. no trigger words.
Trained on Anylora.
Recommended weights 0.5-0.8.
This is my first lora, I don't really understand most of how this works. Just try and see how it works for you.
Description
Resized the original Lora
FAQ
Comments (6)
Can you share your checkpoint?
It's just a mix of sxzluma,aniverse,bbmix,galenaredux and onedoomedmix. I was seeing what those checkpoints were generating and combined them at random according to what I felt was needed to work with queenbee and other art I was training. I don't remember the exact ratios or values. So I didn't want to put it out there when I can't add those details. I will update soon.
nice work! Where do you get the dataset from?
Just manhwa websites,
The best!!