<p>使用vpred+ztsnr+debias estimation loss+scale vpred loss like noise pred+keep tokens separator训练，开启shuffle tokens</p><p>推荐训练模型和推理模型的tag使用顺序为先画风artist name，模拟风格比如jojo (style)和次级风格比如greyscale，然后人物名，服装名，接着是1girl，solo到视角，动作，外貌衣服细节，然后是系列版权名copyright，官方，模拟艺术official art或者official style类时间year xxxx周期newest质量词masterpiece，分辨率absurdes，敏感度explicit。结尾lora模型（即画面占比越大，提示词范围越小越精确的词权重排前，画面占比越小越后，概念越混乱越后。以画面占比优先，提示词精确程度在其后考虑）</p><p></p><h1>模型可能过拟合，记得降权使用</h1><h1 id="lorahttps:xypher7.github.iolora-metadata-viewer">lora参数查看器<a target="_blank" rel="ugc" href="https://xypher7.github.io/lora-metadata-viewer/">https://xypher7.github.io/lora-metadata-viewer/</a></h1><p></p><h1 id="https:civitai.comarticles4560upscaling-images-using-multidiffusion">推荐使用<a target="_blank" rel="ugc" href="https://civarchive.com/articles/4560/upscaling-images-using-multidiffusion">https://civarchive.com/articles/4560/upscaling-images-using-multidiffusion</a>方法放大</h1><h1 id="pixel-artnearest4x-animesharpadetailer">或者使用高清修复（因为是pixel art，所以推荐使用nearest缩放而不是4x-animesharp），并配合adetailer插件修脸，否则画面将模糊</h1><p></p><p></p><h2 id="please-support-the-game-winter-memories-thank-you-meow"><span style="color:rgb(193, 194, 197)">（请支持冬日狂想曲游戏谢谢喵/Please support the game Winter Memories thank you meow）</span></h2><p></p><h1 id="https:huggingface.codatasetstlfzwinter_memoriestreemain"><span style="color:rgb(193, 194, 197)">训练文件上传</span><a target="_blank" rel="ugc" href="https://huggingface.co/datasets/TLFZ/winter_memories/tree/main">https://huggingface.co/datasets/TLFZ/winter_memories/tree/main</a></h1><p></p>

winter memories | 冬日狂想曲 | あまえんぼ冬

Files with SHA256: 07b0a3cdb74e95eae7cfee8bcaaefe931f042297f0a2445d0cbabba49298382f