🎞 LoRA is trained purely on anime screencaps
🗝 Activation tags are in the right panel
📸 Always use ADetailer to get detailed eyes
🛠 Technical notes:
LoRA shoud be compatible with Illu models & NoobAI. (Not tested.)
LoRA is compatible with various styles - Lower the weight & remove "anime screencap" from the prompt, if you are getting too strong anime style
Hi. res fix used on all sample images:
Upscaler: R-ESRGAN 4x+ Anime6B
Steps: 10-15
Denoising: 0.3
📬Leave a like, comment, review but first of all, ENJOY ;)
Description
FAQ
Comments (8)
Hey just curious, but illustrious doesn't seem to use trigger words alone, does it always need supporting prompts along with a trigger word to activate the lora?
There are no trigger words on most of my illu loras. They are not necessary. All you have to do is to define the character to use the Lora or simply copy paste what I put into the activation/"trigger" words in the right panel.
@LittleJelly Is that pretty common for illust models? I've noticed that when I train mine they don't activate with just the trigger word like some of my pony models I've trained before. I was thinking maybe I was training them wrong? O_o
Your models always turn out so good too so that's why I wanted to get feedback from someone that knew what they were doing :3
@LindezaBlue ping me on my discord and we can check your settings... also "chara" triggers can absorb some random stuff from the dataset so it's recommended to not use them unless you are training multi-outfit lora with 2 similar outfits
@LittleJelly will do, just got all situated with roles etc in your server. :3
nana plis
Sorry, I'm usually not going back to the series I've released already.
@LittleJelly ok...thank you