DESCRIPTION
Had another go at Stable Diffusion, this time creating a Honoka model via Dreambooth Lora. The engine had a real hard time doing a side ponytail. Out of literally hundreds of renders, I got like, five with it correct.
Anyway, attached is a small handful of the ones I've shopped (eyes are a pain to fix too), along with the models I trained if anyone's interested in trying it out.
To note: with the V1 model, the trigger prompt is "doahonoka", and with the V2, the trigger is "honokadoa". "Sidetail" is also *mean to work :D
I'll have a go at perfecting the next model too