The main purpose of the model is to reduce oiliness and reduce extreme white light. Reduce details that look too professional. Or it looks like an AI generator. The resulting images will be same level as fan art to reaching mainstream anime but not reaching the level of professional drawings CG. This model is strong in terms of generating images of characters in their teenage years.
V1.1 it's week When the generation of things that are black You need to use Lora to adjust the lighting and add additional details.
main by < anmitsumix > add Increase training with animated images of teenage girls.
add Brightness Adjustment lora clearcolor lora R2W_R lora
V2 has braked Vae_klF8Anime2 into the model, added hyperrefiner0.8, added BgDown_flatBG0.5, added RedFace_hohoaka0.3Color_saturation 0.3.
The result is finer, sharper lines. Eyes are more beautiful, black is less dark.
For people who have already downloaded V1, you can use Lora. Adjust the value and use Vae to get the same result as V2.
If you don't like backgrounds that are too rough, use Lora BgDown_flatBG-5 or use Lora Add Detail to help.