This LoRA is about ニュイ・ソシエール(Nui Sociere), a vtuber of Nijisanji.
The training data comes from some Nui Sociere images of artist Fan no hitori. Almost all images is monochrome, so support add prompts "monochrome, greyscale"(If you need color images, I also recommend using Nui Sociere ニュイ・ソシエール / Nijisanji にじさんじ).
This model need more NSFW test.
Maybe you can use this model generate other vtuber, but effect isn't guaranteed.
Description
As a style model, I believe it has failed. Perhaps there are also reasons why Anima finds it difficult to retain the artist style?