<h1 id="premium-character-loras">Premium character LoRAs</h1>Original synthetic characters, bundles and AI tools are in my shop:<a target="_blank" rel="ugc" href="https://rkrra.com/civitai">https://rkrra.com/civitai</a>Get the character LoRAs on Patreon: <a target="_blank" rel="ugc" href="https://www.patreon.com/cw/RkRRa">https://www.patreon.com/cw/RkRRa</a>Join the community on Discord: <a target="_blank" rel="ugc" href="https://discord.gg/nB79wRNtMT">https://discord.gg/nB79wRNtMT</a><h2>Vulcan Female 2</h2>A character LoRA for Krea 2. Reproduces a consistent subject across poses, framing, and lighting.<h3>Usage</h3>Include the trigger word <code>vlcnfml2</code> in your prompt.Recommended weight: 0.7. At weight 1 this LoRA also pulls other people in the image towards this character: ask for someone else and you may still get this face. At 0.7 the likeness holds and the rest of your prompt is respected. If you need a stronger likeness, go up in small steps.Two character LoRAs at once tend to blend, especially when the two characters look alike. For two characters in one image, generate them one at a time.<h3>Prompt Examples</h3>Here are some prompt examples:<blockquote>"vlcnfml2, close-up portrait, eye-level, head slightly tilted, looking off-camera, neutral expression, dark knit sweater, marina at sunset, warm golden light reflecting on water and boats."</blockquote><blockquote>"vlcnfml2, three-quarter from mid-thigh up, slightly low angle, seated relaxed on bench, calm gaze, striped shirt, jeans, tulip field with windmill, bright natural daylight."</blockquote><blockquote>"vlcnfml2, chest-up, slightly low angle, head tilted down, serene expression, urban balcony backdrop, soft diffused daylight, cityscape behind, railing visible, greenery nearby, overcast sky, architectural details in distance."</blockquote><blockquote>"vlcnfml2, close-up portrait, eye-level, turned head gazing forward, neutral expression, black puffer jacket, beanie, dense forest path, dappled sunlight filtering through canopy."</blockquote><blockquote>"vlcnfml2, chest-up, slightly low angle, hands on hips, slight smile, garage with cars, bright overhead lighting, metallic surfaces, modern tools, reflective floor, sleek vehicle in background."</blockquote><blockquote>"vlcnfml2, three-quarter from mid-thigh up, slightly low angle, head tilted back, hands clasped, smiling, floral dress with lace cardigan, stone alleyway with lanterns, warm ambient glow from windows and streetlight."</blockquote><blockquote>"vlcnfml2, close-up portrait, slightly low angle, head turned right, gazing downward, neutral expression, urban night scene with illuminated building facade, warm ambient glow from streetlights and building lights."</blockquote><blockquote>"vlcnfml2, chest-up, slightly low angle, hand near neck, head tilted back, smiling, hallway with carpet and doors, warm ambient lighting from wall sconces, soft shadows, elegant decor."</blockquote><blockquote>"vlcnfml2, chest-up, eye-level, facing slightly right, head tilted, neutral expression, outdoor market stalls with produce, overcast sky, soft diffused lighting, people in background, wooden crates, hanging greenery, stone pavement."</blockquote><blockquote>"vlcnfml2, chest-up, eye-level, standing, gazing right, wearing jacket and tank top, coastal rocky beach backdrop, soft diffused daylight."</blockquote><blockquote>"vlcnfml2, full length from behind, camera level, standing with hands by sides, head turned slightly right, dimly lit bar with shelves, warm ambient glow from lamps and bottles, polished floor reflecting soft light."</blockquote><blockquote>"vlcnfml2, three-quarter from mid-thigh up, slightly low angle, seated cross-legged, hands clasped, serene smile, olive outdoor gear, terraced rice fields and mountains, golden hour backlighting."</blockquote><blockquote>"vlcnfml2, three-quarter from mid-thigh up, slightly elevated angle, seated on table, legs crossed, mouth open, surprised expression, solid pink backdrop, soft even lighting, minimal shadows, clean studio environment."</blockquote><blockquote>"vlcnfml2, full length, straight-on, standing with one leg forward, neutral expression, plain brown backdrop, soft directional lighting from front-left, casting subtle shadow behind."</blockquote><blockquote>"vlcnfml2, three-quarter from mid-thigh up, slightly low angle, hand near mouth, holding camera, surprised expression, underwater tunnel with fish, bright ambient light from above, reflective floor, curved glass walls, aquatic environment."</blockquote><blockquote>"vlcnfml2, chest-up, slightly low angle, hands on hips, mouth open mid-speech, confident stance, snowy mountain village backdrop, golden hour sunlight casting warm glow across scene."</blockquote><blockquote>"vlcnfml2, close-up portrait, side angle, head tilted back, smiling, cream knit sweater, seaside wooden pier at sunset, warm golden light reflecting on water, tranquil moment."</blockquote><blockquote>"vlcnfml2, chest-up, eye-level, standing, gazing right, neutral expression, gym with boxing ring, bright natural light from windows, metallic texture visible, urban skyline faintly seen through glass."</blockquote><blockquote>"vlcnfml2, three-quarter from mid-thigh up, slightly low angle, seated on stone ledge, legs crossed, hand resting beside, neutral gaze, urban café patio, bright daylight, shadows cast, potted plants, tables, chairs."</blockquote><blockquote>"vlcnfml2, three-quarter from mid-thigh up, slightly low angle, arms extended gripping bookshelves, head tilted upward, neutral expression, bookstore with wooden shelves and hanging lights, warm ambient lighting from ceiling fixtures."</blockquote><h3>Training</h3><ul><li>Base model: Krea 2 (trained on Raw)</li><li>Type: Character LoRA</li><li>Dataset: 27 images</li><li>Steps: 2000</li><li>Network rank / alpha: 32 / 32</li><li>Learning rate: 0.0002</li><li>Optimizer: adamw8bit</li><li>Resolutions: 512 / 1024</li></ul><h3>Generation settings</h3>Trained on Krea 2 Raw, and transfers to Krea 2 Turbo for faster generation.<ul><li>Raw: ~30 steps, guidance 3.5</li><li>Turbo: ~8 steps, guidance 1.0</li></ul>The preview images were generated with <code>krea2_turbo_mxfp8.safetensors</code>, at 8 steps, guidance 1, LoRA weight 1.

Files with SHA256: 778362a5573dc59fe5ae4f7d8d7ddfd1df93c77e7af71cd79ab2dbdc6eff71f5