<h3 id="heading-726">Introducing a better way to use this LoRA with refiner. All sample images are generated in this way:</h3>Settings:<ol><li>Check "Hires fix: show hires checkpoint and sampler selection".</li><li>Set "Hires fix: which pass to enable refiner for" to first pass".</li><li>Merge LoRA into your base model. You may use SuperMergage <a target="_blank" rel="ugc" href="https://github.com/hako-mikan/sd-webui-supermerger">https://github.com/hako-mikan/sd-webui-supermerger</a></li></ol>Instruction:<ol><li>Set base checkpoint to the merged one in Settings#3.</li><li>Set Hires, refiner, ADetailer (optional) checkpoint to the one with your desired style (For example, I merged my SDXL style LoRA into as Animagine XL V3 a style checkpoint)</li><li>Suggested configurations: Sampling steps 28Refiner switch at 0.425</li></ol>基于refiner的模型使用手法 <a target="_blank" rel="ugc" href="https://www.bilibili.com/video/BV1rN4y1B7eg/">https://www.bilibili.com/video/BV1rN4y1B7eg/</a>前期设置：<ul><li>在设置中开启高清修复模型选择</li><li>在设置中把refiner改为first pass,这样确保在高清修复的过程中不使用refiner</li><li>把立绘法LoRA融进你的基础模型中</li><li>融合插件推荐使用: SuperMergage <a target="_blank" rel="ugc" href="https://github.com/hako-mikan/sd-webui-supermerger">https://github.com/hako-mikan/sd-webui-supermerger</a></li><li>（可选）推荐使用ADetailer插件增加面部细节：<a target="_blank" rel="ugc" href="https://github.com/Bing-su/adetailer">https://github.com/Bing-su/adetailer</a></li></ul>使用设置：<ol><li>把基础模型设置为融合了立绘法LoRA的模型</li><li>把高清修复和Refiner模型设置为你想用的风格化模型</li><li>根据我的测试，迭代步数28，refiner切入时机设置为0.425</li><li>如果你使用了ADetailer插件，把ADetailer的inpaint模型也改成同样的风格化模型</li></ol>

Files with SHA256: 1ac9f678d21e1eee705bcc8e4d64e28335c6f873ff3bd8c9bb901ccbfe9a8795