-- created Pony v3 with an updated dataset, threw in an SDXL version while baking that with the same source. Haven't fully tested SDXL with SDXL prompting since I didn't change the data tagging approach specifically for it.. So ,Pony prompting may work better but it did seem to work well in the limited testing I did perform --

With the buzz of his some of previous content out there in the wild, I thought I'd bake a Luigi luigi on Pony using my own workflow using high quality images in the dataset. I thought the results were fairly good for a relatively quick effort.

With this version, ADetailer highly recommended for wide/full shots since this dataset run didn't include that many of that type.

Deny. Defend. Diffusion.

main trigger word: luigiman