Trained on 273 pics with various characters and positions. I think it captured the style quite well.
nanod.safetensors