My first official Style Lora - still in Testphase. Feedback would be great :)

It can have an huge impact - try around with the weighting.

It was trained with sfw images mostly. For ???? Images you may adjust your prompt.

Most of the Preview images are made with prompts from 3Humps1Kiss :)

Triggerword is kinda optional for now. Try with and without :)

The Lora is trained on my own images from Midjourney Niji. So far it gives good eyes and shadowing, smoothing the images and enhance effects. Still need to test more ^^

Breaks at Realism Checkpoints. But you can try.

Works at around 0.5 - 1.0 weight. It depends on the ckp :)

CFG: 3-5

Sampler: Euler a

Should work with other Loras - maybe try different weightings.

I don't create at CivitAI on my own... so please tell me, if something is off