Version 3 is here, trained using existing knowledge, activation tags have been changed

"Hellhound, monster girl encyclopedia, black skin, black sclera, red eyes,"

90% of the time this will generate her accurately with animal hands and feet, feel free to add those tags and "wolf tail" if it fails to show up or you require more precision.

is capable of some 3d renders, though I am not a fan of it so I didn't test much of that aspect.

Eyes will require inpainting since they are kind of rough, eye flames are toggleable to an extent, using negative prompt, sometimes wierdness will appear even if you leave them off since this uses the base models knowledge. So I prefer to leave them off and in the negative prompt. All in all I am very happy with how Version 3 turned out, learned a lot in the process through lots of trial and error.

Feel free to post some stuff you make I want to see it, thanks and enjoy.

[Version 2] outdated, Increased Images in Training Data Set, Retrained On Anime PonyXL to Eliminate Non Anime influence on the model, seems to have fixed the random anthro/furrys showing up in generation since Anime is the focus on this model, Ms Hellwan is much more consistent now. Recommend CFG 6-7 and any of the AutismMix models.

[Version 1 Obsolete], will leave available as Legacy,Created and Trained from Pony Diffusion V6 XL. I have tried it on other models and the results were less than satisfactory, Trigger word is 'Hellwan' I also use "black skin, colored skin, animal hands, animal paws, black sclera, anime/source_anime, to get closer to accurate on model generations, will still sometimes produce anthros/furrys even with negative prompts. Many thanks to the creator KC and the fan artists who created the images this is trained on. The first lora I have created that I am satisfied with the results but I do aim to improve upon it in the future