Inspired by a recent NAI v3 gen of Anon on 4chan and after failing miserably to replicate it in Pony (probably a skill issue) I decided that some cooking had to be done. I didn't use realistic images in the dataset but in my tests it seems to still work on the Real Pony Checkpoint.
Although v7 has the data to generate the male on female piggyback concept the probability is low. This version has only images with a male sitting on the back of a female. I was able to find only 16 decent images for the dataset so this LoRA isn't very flexible but may still be usable. It also has difficulties generating good eyes.
Example Triggers:
piggyback riding, sitting on person, , , , human chair, 1boy, 1girl, from side, profile
holding leash, , gagged
Just sitting:
human chair, sitting on person, , , 1boy, 1girl
crossed legs
Trained perspectives (some work better than others):
from side
straight-on
from above
from below
Trained but lower probability to trigger:
boy on top
Trained only on images where the male is sitting on the back of a female. Due to a low number of images in the dataset (16) it isn't very flexible. It also has difficulties generating eyes.
1. Dieses Modell dient nur Lernzwecken. Urheber- und Auslegungsrechte liegen beim Originalautor.
2. Bist du der Originalautor eines Modells, kontaktiere uns bitte zur Authentifizierung über unsere offiziellen Kanäle. Wir schützen die Rechte aller Schöpfer. Hier klicken, um es zu verifizieren.
