This is basically my first attempt at creating a LoRA for a style. His female counterpart was easily available, but I had problems with him.
Although I trained with 768×768px images, you should get good looking pictures from 512×512px on up with the most basic/standard settings (Euler A, 20 Steps, CFG 7). I tested with Checkpoints that can create photorealistic images, and all worked well.
No trigger word needed, works also with celebrities and women; be aware, though, that I decided against doll hair in the training images (it does show up sometimes).
"Close up" and "full body shot" were used in the descriptive texts for the training. Might need highres fix for full body shots.