A combined style / detailer Lora for SD 1.5, trained at native 1024px on the same dataset of 1000 Flux gens as Zoot Fluxxer XL, all captioned with both Florence-2 Large "more detailed mode" natural language descriptions and also Booru tags from wd-eva02-large-tagger-v3. No particular trigger word. Recommended strength is around 0.7 to 0.8ish, but feel free to adjust as you see fit.

Read the "About This Version" section of the page for more detailed info on any given specific version of the Lora.

All initial showcase images were generated with just Base SD 1.5 and this Lora, at a minimum of 768x1024 / 1024x768 initial resolution before any upscaling.

This Lora is not useful for generations at regular 512px-range SD 1.5 resolutions, use it for hi-res-fix or to help with generating directly at higher resolutions.