Hi everyone!

This is the first version, but it is already usable for generations.

Parameters I used for the generations

- Lighting LoRA: 4 steps

- Steps: 4–8

- Sampler: Euler A

- Scheduler: beta57

This LoRA is not meant to be completely photorealistic.

The main goal is to explore and unlock things that were not possible with the original model.

The training set included 500 images for the first try, and none of them were taken from real people.

That is what made it possible to achieve such unusual body proportions.