Lee-Ahn from the adult webtoon Happy Campus

No POV for this one, instead there's around 10 images each for a variety of checkpoints. Each batch of ~10 is split further into two slightly different prompts that tend to result in a more stylised look in the first half and semi-real(ish) in the second half.

Hopefully these examples show how well the LoRA performs on each checkpoint. Performance on different checkpoints was one of the main issues with the Sweet Guy set, this should be much improved - with the exception of base ponyDiffusion for some reason, things stay weird there.