The Chunin + The Last Hinata LoRA refines Hinata Hyuga’s mature appearance for SDXL models that already recognize her. Trained with increased network capacity (26/13) and low learning rates, it captures both Chunin and The Last era outfits while preserving facial likeness and natural proportions. Careful step scaling and consistent captions allow the LoRA to enhance era-specific details without overwriting the base model’s priors, ensuring stable, accurate results across poses and seeds.