This is a high-fidelity SDXL LoRA trained to generate Ogino Chihiro from the movie Spirited Away in the iconic Studio Ghibli art style.This model was the winner of a rigorous comparison test against another LoRA trained on the Illustrious XL base model. It consistently produced higher CLIP scores and was qualitatively judged to have superior character consistency and style adherence. Trigger Words &amp; UsageTo use this model, you must include the following in your prompt: ogino_chihiro, 1girl<ul><li>ogino_chihiro is the main trigger word for the character.</li><li>1girl is included for better stability and compatibility with the Animagine XL base model.</li><li>solo can also be added for single character focus.</li></ul>Recommended Settings:<ul><li>Base Model: Animagine XL V3.1 (<a target="_blank" rel="ugc" href="https://huggingface.co/cagliostrolab/animagine-xl-3.1">https://huggingface.co/cagliostrolab/animagine-xl-3.1</a>)</li><li>LoRA Weight: 0.8 (can be adjusted between 0.7-0.9 for desired strength)</li><li>CFG Scale: 6.5</li><li>Steps: 30</li><li>Sampler: Euler a (Euler Ancestral Discrete Scheduler) or DPM++ 2M Karras</li><li>Resolution: 1024x1024</li></ul>Prompting Advice:<ul><li>The model was trained extensively using Danbooru-style tags. For best results, use descriptive tags for features, clothing, expressions, and poses (e.g., brown_hair, ponytail, pink_robe, surprised_expression, full_body).</li><li>To emphasize the Ghibli aesthetic, include ghibli_style in your positive prompt.</li><li>For enhanced quality, use tags like masterpiece, best quality.</li><li>For detailed eyes, consider adding large_eyes, shiny_eyes.</li><li>Using negative prompts is highly recommended. A good starting point is: lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, jpeg artifacts, signature, watermark, username, blurry, adult, photorealistic, nsfw</li></ul>

Ogino Chihiro (Ghibli Style) - Animagine XL v3.1

Initial release of the Ogino Chihiro (Ghibli Style) LoRA. This version is the result of extensive testing and optimization, demonstrating superior character consistency and style adherence when used with Animagine XL 3.1 as the base model.

Files with SHA256: 26771815e6e0af0e3e61bdd31f088ecfa5c6c2dd78e17e0af4d879462c245938