Character LoRA trained on photos of a real person.

Best results are achieved with lower LoRA weights (0.6–0.75).

Designed for realistic portrait photography, YouTube thumbnails, and casual photo-style generations.

Recommended settings:

– LoRA weight: 0.6–0.75

– CFG: 4.5–6

– Avoid beauty, fashion, or stylized prompts

– Use natural lighting and realistic camera prompts

This LoRA may produce exaggerated facial features at higher weights.

For best likeness, use earlier training epochs (e.g. epoch 10–11).