<h3>Follow Shano Pirika</h3><ul><li><a target="_blank" rel="ugc" href="https://www.pixiv.net/users/20164912">https://www.pixiv.net/users/20164912</a></li><li><a target="_blank" rel="ugc" href="https://twitter.com/Shanopirika">https://twitter.com/Shanopirika</a></li><li><a target="_blank" rel="ugc" href="https://www.instagram.com/shanopirika/">https://www.instagram.com/shanopirika/</a></li></ul><h3>ShanoPirika Style Lora.</h3><ul><li>Trained with 50 pictures, the basic model is MIX-Pro-V3, you can use the same model to get better result. You can try another model. I had only tried with the MIX-Pro-V3.</li><li>Suggest weight is around 0.9, and maybe it's hard to generate complex background</li></ul><h3>Example</h3>you can see the example pictures.<h3>关注 Shano Pirika</h3><ul><li><a target="_blank" rel="ugc" href="https://www.pixiv.net/users/20164912">https://www.pixiv.net/users/20164912</a></li><li><a target="_blank" rel="ugc" href="https://twitter.com/Shanopirika">https://twitter.com/Shanopirika</a></li><li><a target="_blank" rel="ugc" href="https://www.instagram.com/shanopirika/">https://www.instagram.com/shanopirika/</a></li></ul><h3>ShanoPirika 风格 Lora.</h3><ul><li>由50张图片训练而来，基础模型使用了MIX-Pro-V3，为了得到更好的结果，建议同样使用MIX-Pro-V3，我只使用该模型进行了测试。</li><li>建议权重为0.9左右，另外可能难以生成复杂的背景。</li></ul>This is Tag-Frequency references:以下是训练集tag频率参考图：<img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/d8faeccb-c97d-45c6-fdbd-95da098ccf00/width=525/d8faeccb-c97d-45c6-fdbd-95da098ccf00" />

ShanoPirika Style

ShanoPirika Style

Files with SHA256: f35601a03a27faf2ddcd3cdba007eddea831b7c03a3e7d2804ebcae62ffa4a69