SDXL LoRA trained with Civit-AI trainer for Pony V6/Pony merges. This is an experimental kind of LoRA, aimed with the goal of improving understanding of the finer details of fur and light when it comes to furry-type subjects. Can also be used with human subjects, but note the use of e621 tagging.

No triggerwords, uses e621 styled prompting. Recommend Adetailer, or negative hand embeddings might help with consistence of fingers.

Optional quality keywords to add might be: absurd res, hi res, digital drawing (artwork), digital media (artwork), backlighting, countershading, detailed background, detailed fluffy fur