<h1 id="aubrey-lenox">Aubrey Lenox</h1><h3 id="trigger">Trigger</h3><pre><code>blonde-87
</code></pre><hr /><h1 id="description">Description</h1>Aubrey Lenox is a photorealistic Pony character LoRA featuring a naturally beautiful young blonde woman with striking blue eyes, long center-parted blonde hair, defined brows, soft youthful facial features, and a warm, approachable expression.This LoRA was trained on a carefully curated dataset emphasizing identity consistency across a wide variety of poses, lighting conditions, hairstyles, clothing, and environments. The result is a versatile character that performs well in studio portraits, lifestyle photography, glamour editorials, outdoor scenes, and casual everyday settings while maintaining a consistent appearance.Aubrey excels at producing clean, realistic facial features with expressive blue eyes, natural skin texture, and believable lighting. She works especially well for portrait photography, fashion, travel, lifestyle, and cinematic scenes.<hr /><h1 id="recommended-settings">Recommended Settings</h1>Checkpoint<ul><li>Realism by Stable Yogi (recommended)</li><li>Compatible with many Pony checkpoints, though appearance may vary slightly depending on the model.</li></ul>Sampler<pre><code>dpmpp_2m
</code></pre>Steps<pre><code>30–35
</code></pre>CFG<pre><code>4
</code></pre>LoRA Weight<pre><code>0.85–0.95
</code></pre>Recommended starting point:<pre><code>Weight: 0.90
CFG: 4
Steps: 32
Sampler: dpmpp_2m
</code></pre><hr /><h1 id="prompt-tips">Prompt Tips</h1>Use the trigger near the beginning of the prompt.Example:<pre><code>&lt;lora:blonde-87 (aubrey) PONY:0.9&gt;, blonde-87
</code></pre>For the best identity preservation, keep the prompt identity-first, then describe the environment afterward.Good facial descriptors include:<ul><li>gentle closed-mouth smile</li><li>relaxed expression</li><li>direct natural gaze</li><li>symmetrical eyes</li><li>realistic eye anatomy</li><li>detailed irises</li><li>realistic skin texture</li></ul>Avoid overloading the prompt with excessive beauty descriptors, as they can allow the checkpoint to override the learned identity.<hr /><h1 id="example-prompt">Example Prompt</h1><pre><code>&lt;lora:blonde-87 (aubrey) PONY:0.9&gt;, blonde-87, adult woman, long center-parted blonde hair, blue eyes, gentle closed-mouth smile, relaxed expression, direct natural gaze, realistic skin texture, standing beside the ocean at sunset wearing a light blue summer dress, soft golden-hour lighting, photorealistic
</code></pre><hr /><h1 id="example-negative-prompt">Example Negative Prompt</h1><pre><code>multiple people, duplicate person, different face, generic blonde, child, teen, male, short hair, curly hair, bangs, heavy makeup, asymmetrical eyes, lazy eye, wandering eyes, crossed eyes, mismatched pupils, malformed iris, bad anatomy, bad hands, blurry, cartoon, anime, illustration, painting, plastic skin
</code></pre><hr /><h1 id="works-great-for">Works Great For</h1><ul><li>Photorealistic portraits</li><li>Lifestyle photography</li><li>Fashion photography</li><li>Glamour portraits</li><li>Outdoor photography</li><li>Beach scenes</li><li>Editorial photography</li><li>Luxury environments</li><li>Professional business portraits</li><li>Coffee shops and cafés</li><li>Travel photography</li><li>Casual everyday scenes</li></ul><hr /><h1 id="training-notes">Training Notes</h1>This LoRA was trained using a carefully curated dataset focused on:<ul><li>consistent facial identity</li><li>varied lighting</li><li>multiple hairstyles</li><li>diverse environments</li><li>natural expressions</li><li>portrait and environmental compositions</li></ul>The goal was to create a flexible character LoRA that maintains Aubrey's identity across a broad range of prompts rather than overfitting to a single pose or style.Enjoy creating with Aubrey Lenox!

Files with SHA256: 0c82060428a7850f893a4bf9e631f107b6b2af85600c449d79070f4213a61a2e