LoRA trained in the style of "OkiOppai".
Removed watermark from dataset.
OkiOppai_Style_Illust.safetensors