<h2>M71z30 3DCG Style (MOSS-Image)</h2>
<p>Seamless Cel-Shaded Anime Contours Fused with Next-Gen 3D Shading</p>
<p>Trigger Word: <code>style_m71z30</code>  |  Recommended Weight: <code>1.0</code></p>
<p>Trained exclusively on MOSS Image Base at native 1024px resolution with constant LR <code>0.00020</code>, Rank <code>32</code>, Alpha <code>48.0</code>, and Pure Style Baking.</p>
<hr />
<h3>🌟 Key Visual Characteristics &amp; Strengths</h3>
<ul>
<li>Cel-Shaded &amp; 3D Synergy: Clean anime line art boundaries blended with smooth 3D gradient shading.</li><li>Vivid Expressive Faces: Dynamic facial expressions, perfectly symmetrical eyes, and lively anime proportions.</li><li>Saturated Pop Palette: Rich color vibrancy that maintains clarity even in complex ambient lighting.</li><li>Dynamic Character Posing: High prompt responsiveness to kinetic motion, perspective foreshortening, and iconic poses.</li>
</ul>
<hr />
<h3>💡 Pure Style Baking (Zero Prompt Conflicts)</h3>
<p>During dataset curation, all character clothing, anatomy tags, and generic lighting/render keywords (such as <code>3d</code>, <code>cgi</code>, <code>blender</code>, <code>shiny skin</code>) were strictly separated and filtered. The underlying 3D volumetric structure, shading geometry, and subsurface scattering (SSS) are deeply baked into the LoRA weights.</p>
<p>Benefit: You have complete creative freedom over character outfits, poses, lighting setups, and atmospheres without model collapse or visual fighting.</p>
<hr />
<h3>📝 Prompting Examples</h3>
Character Portrait:
<pre><code>best quality, highres, style_m71z30, 1girl, solo, playful wink, colorful hoodie, bright street lighting, dynamic perspective</code></pre>
Cinematic Atmosphere:
<pre><code>best quality, highres, style_m71z30, 1girl, standing in futuristic neon alley, vibrant colors, detailed anime face, cinematic neon reflections</code></pre>
Recommended Negative Prompt:
<pre><code>worst quality, low quality, bad anatomy, deformed hands, missing fingers, extra limbs, watermark, text, signature, blurry, oversaturated</code></pre>

M71z30 3DCG Style (MOSS-Image)

<p>Seamless fusion of cel-shaded anime contours and next-gen 3D shading, expressive facial features, and vibrant saturated tones.</p><p>Optimal fitting point at Step 364.</p>

Files with SHA256: e475d90d4b204878f33d2a8754f387899734b99292e9772bbd68e98b8f0ad467