This is a LoRA made with 1 AI generated picture (you can see the sample attached). It should work nicely with portraits and close-ups.