<p>本模型有三个国家的语言介绍</p><p>中文，英语，日文</p><p>This model has language introductions for three countries</p><p>Chinese, English, Japanese</p><p>このモデルには3つの国の言語が紹介されています</p><p>中国語、英語、日本語</p><h2 id="-0cbbqb113"><strong>===================================</strong></h2><h2 id="bv8.0novalai40.-01c557x59"><strong><span>中文：β版本（v8.0）基于NovalAI的原模型使用了40万张图片进行了训练，希望为大家带来一个有关于可爱风格的特化模型.</span></strong></h2><h2 id=":-pingd1h6f"><strong>下面是我对于模型训练的初衷与其他模型的注意事项:</strong></h2><h2 id="-7olq7ztvn"></h2><h2 id="ai-bptm5id3f"><strong>初衷：希望可以探索到绘画ai的极致潜力</strong></h2><h2 id="-g35ysunyu"></h2><h2 id="-obw88aort"><strong>模型注意事项：</strong></h2><h2 id="1.hires.fix-ydix5r9tx"><strong>1.使用Hires.fix可以给予模型更优秀的输出效果。</strong></h2><h2 id="2.sd1.5lora-yrzgsqqfi"><strong>2.适配很多基于SD1.5的lora，并可以有十分优秀的混合效果。</strong></h2><h2 id="-fuutq6rhf"><strong>—————————————————————————</strong></h2><h2 id="chinese:-the-bv8.0-version-is-based-on-the-original-novalai-model-which-was-trained-using-400000-images.-we-hope-to-bring-you-a-specialized-model-about-cute-styles-y8xvbckyr"><strong>Chinese: The β（v8.0） version is based on the original NovalAI model, which was trained using 400000 images. We hope to bring you a specialized model about cute styles</strong></h2><h2 id="-djpmvok5e"></h2><h2 id="here-are-my-initial-intentions-for-model-training-and-considerations-for-other-models:-phrtd4dts"><strong>Here are my initial intentions for model training and considerations for other models:</strong></h2><h2 id="-hmkm3r4rn"></h2><h2 id="original-intention:-i-hope-to-explore-the-ultimate-potential-of-ai-in-painting-hcn97b23x"><strong>Original intention: I hope to explore the ultimate potential of AI in painting</strong></h2><h2 id="-ppcqjy8rr"></h2><h2 id="model-considerations:-f5u9suceb"><strong>Model considerations:</strong></h2><h2 id="using-hire.fix-can-provide-the-model-with-better-output-performance.-b1ou12nlx"><strong>Using Hire.fix can provide the model with better output performance.</strong></h2><h2 id="2.-it-is-compatible-with-many-sd1.5-based-loras-and-can-have-excellent-mixing-effects.-vi974leem"><strong>2. It is compatible with many SD1.5 based Loras and can have excellent mixing effects.</strong></h2><h2 id="-0g1gze2ww"><strong>—————————————————————————</strong></h2><h2 id="bv8.0novalai40-id93ujrsj"><strong>日文：β（v8.0）バージョンのNovalAIに基づくオリジナルモデルは40万枚の画像を用いて訓練を行い、かわいらしいスタイルに特化したモデルをお届けしたいと考えています。</strong></h2><h2 id="-8zcczlnig"></h2><h2 id="-at1t9w18z"><strong>次に、モデルトレーニングの初心と他のモデルについての注意事項を示します。</strong></h2><h2 id="-8s3viv6ih"></h2><h2 id="ai-x4whma7zy"><strong>初心：絵画aiの究極の潜在力を探求したい</strong></h2><h2 id="-h6pv4vymw"></h2><h2 id="-thbit84x5"><strong>モデルの考慮事項：</strong></h2><h2 id="1.hires.fix-da5dx0ekq"><strong>1.Hires.fixを使用すると、モデルにより優れた出力効果を与えることができます。</strong></h2><h2 id="2.sd-1.5-lsbjvpecr"><a target="_blank" rel="ugc" href="http://2.SD"><strong>2.SD</strong></a><strong> 1.5ベースのローラを多く配合し、優れた混合効果を得ることができます。</strong></h2><h2 id="-s9amf2rri"><strong>===================================</strong></h2>

Files with SHA256: e284f791ce5aa2c2bb0023c10d1a054e6468ce20fcb58409009e4808f8fb35ce