This particular version of the Lora was trained exclusively on portrait images. However, thanks to the powerful learning capabilities of the flux model, the Lora can apply its pencil drawing techniques to a diverse range of artistic subjects beyond just portraits.