This LoRA is trained on Pony Realism v2.3 to boost fur detail, density, and overall quality. It should also work with other realistic checkpoints.