This is a test model.
I am working on significant improvements to every aspect of the model. I tried to make sure that this model was not a model of one pose.
There are very few suitable images, so I added heavily processed images created by artificial intelligence and diluted it all with a modest set of images.
I've done my best to make this model self-sufficient enough to produce different but correct results. Don't use all the trigger words at once! It's useless!
I also tried to simplify the generation of various poses and make the model less chaotic.
I hope that the drops will appear more often in the foreground than in the background!
Feel free to criticize and make recommendations.
Description
FAQ
Comments (9)
awesome work, would love to see a pussy (squirt) version of this haha, awesome :)
What checkpoint did you use for your example images?
I use Pony Diffusion V6 Turbo DPO merge with the most inadequate comfyui workflow. Cuz it is fast.
@No_Rizz_Dude Ah okay. I liked the overall art style of your outputs. That's why I'm asking.
The model doesn't recognize the trigger words.
For some reason, I couldnt get v3.0 to work, no matter what trigger words I used, what checkpoint, or whatever other settings copied from example images.
v2.0 however works like a charm.
Weird.
Trying this mainly using Holy Mix [illustriousXL] and couldnt get v3 or v2 to work with it.
yeah because its a pony lora^^
A version for IL would be amazing