Dark

illustrious and NoobAI love high brightness, even if you prompted something dark. Because anime dataset is very bright and colorful. This LoRA can reduce this bias. And make it more photography-ish.

It's also a bootstrap LoRA for "Stabilizer" and "Touching Grass". Use this LoRA to active their full dark image dataset.

Trained on a sub dataset of "Touching Grass", only low brightness images.

It's dark, not black. The main goal of this LoRA is to reduce the brightness bias (or more bias to dark). The training dataset still has a quite wide brightness range. E.g. cityscape at night and full of small building lights.

Only environment, no human in dataset, so it will not "pollute" your base model style. Can be applied on both pure anime and realistic models.

Useful if you want:

• something very dark

• lower the overall brightness and create a photography feeling

Trained on illustrious v0.1, but I tested it on NoobAI as well. The effects are very good. So I don't think we need a separated NoobAI version.