A LoHa trained on a measly 10 images using my method. 1 batch size.

She will be more anime/human or she will be closer to the original depending on prompt combinations and version used.