<h3 id="personal-recommended-settings-ttn3qbc2d">Personal Recommended Settings</h3>Positive Prompt<pre><code>masterpiece, best quality, very aesthetic, highres, nagi, &lt;lora:nagi_IllustriousXL_v1.0:1&gt;, </code></pre>Negative Prompt<pre><code>worst quality, low quality, worst aesthetic, lowres,bad anatomy, interlocked fingers</code></pre>Sampler<ul><li>Sampling method: Euler a / Euler SMEA (requires extension)</li><li>Schedule type: Automatic</li><li>STEP: 35</li><li>CFG scale: 8</li></ul>Hires. fix<ul><li>Upscaler: Remacri (needs to be added)</li><li>Hires stepsc: 0.45</li><li>Upscale by: 1.5</li><li>Hires CFG Scale: 7</li></ul>ADetailer (requires extension):<ul><li>ADetailer detector: face_<a target="_blank" rel="ugc" href="http://yolov8n.pt">yolov8n.pt</a></li><li>Inpaint mask blur: 6</li><li>Inpaint only masked padding, pixels: 96</li><li>Inpaint denoising strength: 0.2–0.4</li><li>Inpaint: 1024×1024</li><li>ADetailer sampler: Euler a※that ADetailer will not work properly if Euler SMEA is selected as the ADetailer sampler.</li></ul>Extensions<ol><li>advanced_euler_sampler_extension<ul><li>GitHub link<a target="_blank" rel="ugc" href="https://github.com/ananosleep/advanced_euler_sampler_extension.git">https://github.com/ananosleep/advanced_euler_sampler_extension.git</a></li><li>Enables the use of Euler SMEA sampler without dy.</li><li>Unlike Euler a, I feel that almost nothing is added to the output other than what is stated in the prompt.</li></ul></li><li>ADetailer<ul><li>GitHub link<a target="_blank" rel="ugc" href="https://github.com/Bing-su/adetailer.git">https://github.com/Bing-su/adetailer.git</a></li><li>Automatically detects and repairs details in the generated image, such as faces and hands.</li></ul></li></ol>Addition required:<ul><li>Remacri<ul><li>Civitai link<a target="_blank" rel="ugc" href="https://civarchive.com/models/147759/remacri?modelVersionId=164821&amp;form=MG0AV3">https://civarchive.com/models/147759/remacri?modelVersionId=164821&amp;form=MG0AV3</a></li></ul></li></ul><h2 id="if-you-like-this-lora-please-like-comment-or-send-us-an-illustration-you-created!-pgayn8xpr">If you like this Lora, please like, comment, or send us an illustration you created!</h2><h3 id="-vdl7avjbz">個人的なおススメ設定</h3>ポジティブプロンプト<pre><code>masterpiece, best quality, very aesthetic, highres, nagi, &lt;lora:nagi_IllustriousXL_v1.0:1&gt;, </code></pre>ネガティブプロンプト<pre><code>worst quality, low quality, worst aesthetic, lowres,bad anatomy, interlocked fingers</code></pre>サンプラー<ul><li>Sampling method：Euler a / Euler SMEA(拡張機能が必要)</li><li>Schedule type：Automatic</li><li>STEP：35</li><li>CFG scale：8</li></ul>Hires. fix<ul><li>Upscaler：Remacri(追加する必要あり)</li><li>Hires stepsc：0.45</li><li>Upscale by：1.5</li><li>Hires CFG Scale：7</li></ul>ADetailer (拡張機能が必要):<ul><li>ADetailer detector：face_<a target="_blank" rel="ugc" href="http://yolov8n.pt">yolov8n.pt</a></li><li>npaint mask blur：6</li><li>Inpaint only masked padding, pixels：96</li><li>Inpaint denoising strength：0.2–0.4</li><li>inpaint：1024×1024</li><li>ADetailer sampler：Euler a※ADetailerサンプラーにEuler SMEAを選択した場合、ADetailerが正しく動作しないので注意。</li></ul>拡張機能<ol><li>advanced_euler_sampler_extension<ul><li>GitHubリンク<a target="_blank" rel="ugc" href="https://github.com/ananosleep/advanced_euler_sampler_extension.git">https://github.com/ananosleep/advanced_euler_sampler_extension.git</a></li><li>dy抜きのEuler SMEAサンプラーが使える。</li><li>Euler aとは異なり、ほとんどプロンプトに記載されている内容以外を出力時に付け加えない気がします。</li></ul></li><li>ADetailer<ul><li>GitHubリンク<a target="_blank" rel="ugc" href="https://github.com/Bing-su/adetailer.git">https://github.com/Bing-su/adetailer.git</a></li><li>生成された画像の顔や手の崩れを自動的に検出し、詳細を修正するためのマスキングとインペインティングをしてくれる。</li></ul></li></ol>追加が必要:<ul><li>Remacri<ul><li>Civitaiリンク<a target="_blank" rel="ugc" href="https://civarchive.com/models/147759/remacri?modelVersionId=164821&amp;form=MG0AV3">https://civarchive.com/models/147759/remacri?modelVersionId=164821&amp;form=MG0AV3</a></li></ul></li></ul><h2 id="lora-zs2pqef2h">このLoraを気に入っていただけたら、いいねやコメント、作成したイラストを送ってくださると嬉しいです！</h2>

nagi_Style

Through fine-tuning various LoRA parameters, we've discovered the optimal configuration that aligns generated images more closely with the desired art style. This has led to substantial improvements in the accuracy of our results! LoRAのパラメータを複数調整し、最適な組み合わせを見出しました。これにより、生成される画像の画風との一致度が大幅に向上し、より高精度な結果を得られるようになりました！

Files with SHA256: fec5ad9ccb60ffa7be124bb551250b397baa15c2e39508d4bdf8f9da219cfcb5