<h2 id="description">Description</h2>High-quality character LoRA of Revy from Black Lagoon, trained for SDXL illustration models.This LoRA focuses on maintaining clear facial identity, stable head structure, and consistent in-character expressions, aiming for reliable results across different prompts and compositions.It is optimized for Illustrious-based SDXL models, I recommend <a rel="ugc" href="https://civarchive.com/models/376130/nova-anime-xl">Nova Anime XL</a> for better results.<hr /><h2 id="what-this-lora-focuses-on">What this LoRA focuses on</h2><ul><li>Consistent and recognizable facial features.</li><li>Stable head proportions.</li><li>Natural, in-character expressions.</li><li>Reliable performance in SDXL illustration workflows.</li></ul><hr /><h2 id="general-usage-recommendations">General usage recommendations</h2><pre><code>Trigger word: revy_blacklagoon</code></pre>The following values are general reference settings that work well with this LoRA.Users are encouraged to adapt them to their own models, workflows, and project needs.<ul><li>LoRA strength: 0.7 – 1.0</li><li>CFG: 6 – 8</li><li>Steps: 25 – 40</li><li>Sampler: DPM++ 2M or 2M SDE with Karras </li></ul> This is my first LoRA of an anime character, if you notice any mistakes or things to improve you can tell me so I can improve it in future versions.

Revy (Black Lagoon) | SDXL / illustrious | Character LoRA

This is an initial release focused on capturing Revy’s facial identity and overall character presence for SDXL illustration models.

Files with SHA256: 64b533ae3d2815b9a8952ab199b719b9e3467f52c53f1c8e0d089d6bc42a6f4c