A LoRA for producing images of Asian ball-jointed dolls, trained on 30 images.

The strength is fairly versatile; for both the ??dim and 128dim I recommend somewhere between 1.0 and 1.3, with 1.2 being pretty solid. There is no trigger word, but I highly recommend guiding your prompts to include "ball-jointed" somewhere in there for best results. Also, if you want an anime-style face, add "anime face" to your prompt (there are a couple of examples in the sample images).

Like my work? Please consider leaving a tip to help fund future LoRAs or commission a custom LoRA for yourself. Higher quality versions of my LoRAs are also available for a monthly donation: https://ko-fi.com/artfulgenie.