This model is trained on a short few of Panos' incredibly unique streetwear artworks (of which I implore you to check out for yourself!)

This LoRA can be a bit finicky to work with (it's my first), so I recommend using a higher CFG scale than usual (10-15) in order to make the colors pop and make it easier to prompt out any artifacts. It works for both girls and boys, but if you want men you need to put feminine characteristics in the negative prompt. Character LoRAs do work, but you need to weight them quite low to avoid style mixing.

I would only recommend Pony V6 but autismMix can work if you figure out a way to fix the wide eyes (which seemed to be quite an issue with autismMix). Regardless, I'm sure other Pony derivatives work fine as well with some finetuning.

Keywords aren't necessary, but invoking the style can help coherence (streetwear, iridescent glasses, 90s anime style, etc.). Basically, use anything that pops out as applicable to the original style.

I appreciate any feedback, and again, if you haven't ever checked out ssgpanos' socials, I highly implore you to do so.