You can try any model you want, however I really suggest using <a target="_blank" rel="ugc" href="https://civarchive.com/models/376031/hassaku-xl-hentai">Hassaku XL Checkpoint </a>or <a target="_blank" rel="ugc" href="https://civarchive.com/models/288584?modelVersionId=324524">AutismMix checkpoint</a>Note: When using Hassaku you might need to use an <a rel="ugc" href="https://civarchive.com/models/402667/age-slider-lora-or-ponyxl-sdxl">age slider</a> though, I like the quality and the prompt adherence of the model, but it tends to age down the characters. (Or just use any other model of your preference)LoRA settings:<ul><li>Weight: 0.7 - 1.0</li></ul>Base model settings:<ul><li>Steps: &lt;30</li><li>Width: 896px</li><li>Height: 1132px</li><li>Sampler: Euler a</li><li>VAE: None</li><li>CFG Scale: 3 - 5</li><li>Clip skip: 2</li></ul>Upscaler settings:<ul><li>Upscaler: Latent (Nearest-Exact)</li><li>Upscale by: 1.5x</li><li>Denoising strength: 0.55</li></ul>

Itsuki Shu Pony // Ensemble Stars!!

I'm trying something different. I'm baking all the character features inside the trigger word so that it only requires 1 word instead of a whole sentence. Let me know if you like this way more than the other, since I believe it's less flexible but I might be wrong.

itsuki_shu_pony.safetensors

v1.0

You can try any model you want, however I really suggest using <a target="_blank" rel="ugc" href="https://civitai.com/models/376031/hassaku-xl-hentai">Hassaku XL Checkpoint </a>or <a target="_blank" rel="ugc" href="https://civitai.com/models/288584?modelVersionId=324524">AutismMix checkpoint</a>Note: When using Hassaku you might need to use an <a rel="ugc" href="https://civitai.com/models/402667/age-slider-lora-or-ponyxl-sdxl">age slider</a> though, I like the quality and the prompt adherence of the model, but it tends to age down the characters. (Or just use any other model of your preference)LoRA settings:<ul><li>Weight: 0.7 - 1.0</li></ul>Base model settings:<ul><li>Steps: &lt;30</li><li>Width: 896px</li><li>Height: 1132px</li><li>Sampler: Euler a</li><li>VAE: None</li><li>CFG Scale: 3 - 5</li><li>Clip skip: 2</li></ul>Upscaler settings:<ul><li>Upscaler: Latent (Nearest-Exact)</li><li>Upscale by: 1.5x</li><li>Denoising strength: 0.55</li></ul>

&lt;p&gt;I&#39;m trying something different. I&#39;m baking all the character features inside the trigger word so that it only requires 1 word instead of a whole sentence. Let me know if you like this way more than the other, since I believe it&#39;s less flexible but I might be wrong.&lt;/p&gt;

Files with SHA256: f67d2366c10a1b1bf52f5eafe0c56eaedba82f55bdd94e537860e73f3a61d33f