<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

Improved version! Re-selected teacher images and trained with 65 images in animagine-xl-3.0.Changes from v1.0.The atmosphere of the image is less likely to change when used with other Lora.The atmosphere of the image can now be maintained even if "masterpiece" or "best quality" is used for the quality tag.I can now draw details a little more carefully (especially fingers, etc. are less likely to break).改良版！教師画像を再厳選しnimagine-xl-3.0で65枚の画像を使ってトレーニングしました。v1.0からの変更点。他のLoraと併用しても絵の雰囲気が変わりにくくなりました。クオリティータグに「masterpiece」「best quality」を使っても絵の雰囲気を保てるようになりました。細部が少し丁寧に描けるようになりました(特に指などが破綻する確率が減りました)。

Files with SHA256: 9f1b0ae68532c45d00ea8e808ff9a11a8f83c1dcc282b042db9f39a210d1f6d5