Version 0.1 is more stable, but much simpler.

Version 0.2 should be version 1.0, as it was trained with many more images, but it is very unstable and has many image bugs, I would discard it, but I decided to release it as version 0.2 for people to test.

This Lora works with strength between 0.5 to 1.0 (recommended is 0.7)

Clip Skip 1 (realistic and 3D model)

Clip Skip 2 (anime model)

Prompts (cellphone pose, holding a cellphone, holding a phone)

(full body, upper body, cowboy shot, from side, kneeling, sitting, crouched)