A Pony art style Lora trained on an assortment of what I've felt were some of the best Flux gens posted to CivitAI so far, all recaptioned with both Florence-2 Large detailed descriptions as well as Booru tags from wd-eva02-large-tagger-v3. Unique trigger word is `flx$tyl3`. Recommended strength is around 0.7 - 0.8, but feel free to experiment. All primary showcase images were generated with Base Pony and this Lora.
For a regular SDXL Lora trained on the same dataset, check out ZootFluxxer XL instead.
Description
Initial version.
FAQ
Comments (4)
Best suited for photorealistic pony, on anime models similar to semi 3D style.
Yeah most Flux content is more realistic so far, I didn't avoid 2D though. May do a V2 with more images at some point.
Where are the crabs?
I think I don't get the joke TBH lol?