Unrealistic ahegao pose just like manga and SD1.5 loras. Also makes the subject Asian and the whole image less saturated.

A proof of concept for Flux Lora training.

Training Set: 40 stable diffusion generated images that follows this guide.

Training Steps: 1200.

Recommended strength: 1.0 .

No need to use all the trigger words at once, if the prompt is simple, merely "open mouth" is good enough for a try.