>>73561808At first I was just getting garbage generations, but that was mostly growing pains with learning learning how and what things work. Then when I could get "an image", I tried to actually play around with it, but custom nodes weren't downloading correctly and I had to fuck around with python installs and some other stuff. Now I'm getting images that generally look fine, but thing like fingers and other details are pretty much always messed up and I'm not sure where or how to clean them up. It feels bad cancelling an image JUST because it has four fingers and I don't know how to fix that.
Overall, I'm still fucking around and finding out.
Right now though, I'm mostly trying to figure out a way to trigger certain outfits or loras without using magic-password-trigger phrases. does <lora:whatever.safetensors:1.0> actually call a lora? do I need EXACT trigger phrases to pull out specific outfits? or is there some plugin that I can use to keep track of these things?
for funsies, here's a curated image of that ame selfie prompt:
https://files.catbox.moe/2izq0s.pnggeneral workflow is largely stolen from here, but slightly and likely poorly edited.
