This model was created by merging LoRAs and some of my older SusaMix models into the anima Aesthetic model.
Unless you use artist tags, it will generally produce an anime-style cel-shaded look.
I recommend using Danbooru tags rather than natural-language prompts.
For quality tags, masterpiece alone is usually enough.
Recommended prompt: masterpiece
(It is preferable to include a Copyright tag and a Character tag.)
Negative prompts: worst quality, nsfw
CFG: 5
shift: 3~10 (I like 10.)
Description
I accidentally uploaded the model with the name “susamix_anima_aesthetic.safetensors” as is.
Feel free to use it however you like.
モデル名をsusamix_anima_aesthetic.safetensorsのままアップロードしてしまった
好きに使ってください