ユーザーLoRA
rastbary
iwfe_35, shimada_art, watercolor, rough
trained with 100 images.
i kinda recommed using the haruka v2 sdxl model, this most likely fits with this LoRa
add "rough outlines, sketch" to the prompt also