This model is trained with V-Parameterization so it will only work on VPred models like NOOB-AI-XL or Obsession, etc.
Noob-AI-XL EPS models will NOT work properly. The image will be deep-fried.
This is a DoRA so make sure to have your stuff somewhat updated
Remember to check out the artist.
Will remove if requested
You'll have to fight bar censors because I was too lazy to prepare my dataset properly...
Use bar censor in negative and uncensored in positive prompt
Might update in the future since I'll be using this dataset to play around with training settings more to see what works better
Illustrious version:
It's alright but doesn't work properly on WAI-NSFW-illustrious-SDXL or Hassaku XL
Pony version:
Might even be able to create some Arknights characters (maybe)
Seems to work but not as strong as that other LoRA... though that thing according to the metadata is trained on like 12 images with no tags... I'm impressed that it works as well as it does...
Always appreciate seeing your generations too
You can DM me to suggest what to train as well or make me finish a particular one mentioned on my profile
I still don't know what I'm doing with training settings :3