This model is trained a set of images that look like a mix between a digital painting and a marker drawing. Its fun, its simple, its easy to use. I recommend running at around 0.75 strength but its powerful and starts having impacts as low as 0.2 strength. At 1.0 strength you are likely to see issues with hands, arms and legs. This LoRA is trained locally on my own machine for Anima, its trained at 128/128 network dimensions so it is a bit larger than most. This Version, the Booru version, is trained on SDXL style tagging instead of natural language but should still apply to both.