<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 1</h2>A character LoRA for Krea 2. Reproduces a consistent subject across poses, framing, and lighting.<h3>Usage</h3>Include the trigger word <code>vlcnfml1</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>"vlcnfml1, full length, low angle, walking forward, neutral expression, snowy street with parked cars, festive string lights overhead, warm streetlights casting soft glow on snow-covered ground."</blockquote><blockquote>"vlcnfml1, three-quarter from mid-thigh up, straight-on, seated with hands resting on knees, serene expression, navy coat, grey scarf, brown boots, cluttered antique-filled room, warm ambient lighting from lamps and window."</blockquote><blockquote>"vlcnfml1, chest-up, eye-level, seated facing right, neutral expression, plaid shirt, olive pants, lush green foliage, dappled sunlight filtering through leaves."</blockquote><blockquote>"vlcnfml1, three-quarter from mid-thigh up, slightly low angle, standing on steps, one leg lifted, gaze forward, rooftop garden with potted plants, bright daylight, urban skyline visible in distance, clear sky above."</blockquote><blockquote>"vlcnfml1, three-quarter from mid-thigh up, eye-level, seated cross-legged, hands resting on knees, neutral expression, layered jacket and cargo pants, industrial container yard, warm golden hour light."</blockquote><blockquote>"vlcnfml1, close-up portrait, eye-level, facing right, neutral expression, blurred kitchen background, warm ambient lighting, soft focus on surroundings, professional environment, subtle depth of field."</blockquote><blockquote>"vlcnfml1, chest-up, straight-on, standing centered, slight head tilt downward, gentle smile, graffiti-covered alleyway walls, soft diffused daylight filtering through narrow space."</blockquote><blockquote>"vlcnfml1, full length from behind, straight-on, standing still, hands at sides, facing window, office with desks and screens, ambient daylight from large windows, futuristic cityscape visible outside, multiple monitors displaying data."</blockquote><blockquote>"vlcnfml1, waist-up, straight-on, arms crossed, laughing, teal blouse, black leather skirt, solid orange background, even studio lighting."</blockquote><blockquote>"vlcnfml1, three-quarter from mid-thigh up, straight-on, standing still, neutral expression, comic book shelves behind, soft even lighting, focused gaze forward, shelves filled with collectibles, professional demeanor."</blockquote><blockquote>"vlcnfml1, chest-up, eye-level, holding notebook and pen, smiling, olive overalls over gray tee, wooden pier with boats and sea, soft diffused daylight."</blockquote><blockquote>"vlcnfml1, three-quarter from mid-thigh up, straight-on, hands in pockets, neutral expression, plain blue backdrop, soft even lighting, poised stance, calm demeanor."</blockquote><blockquote>"vlcnfml1, chest-up, straight-on, hands on hips, neutral expression, stone bridge over water, ancient buildings, soft golden light filtering through trees, serene yet commanding presence."</blockquote><blockquote>"vlcnfml1, chest-up, slightly low angle, hand raised mid-gesture, smiling with one eye closed, dark gradient backdrop, soft directional lighting from front-left, focused on upper body and gesture."</blockquote><blockquote>"vlcnfml1, three-quarter from mid-thigh up, slightly low angle, standing with one foot forward, looking sideways, neutral expression, empty parking lot at night, cool blue artificial lighting from overhead lamps."</blockquote><blockquote>"vlcnfml1, three-quarter from mid-thigh up, straight-on, arms raised overhead clasped, focused expression, golf course with sand traps and trees, bright natural daylight, clear sky, green grass, outdoor setting."</blockquote><blockquote>"vlcnfml1, chest-up, eye-level, drinking from bottle, calm expression, yoga studio with mats and windows, soft natural light from side, wooden walls, plants, people seated in background."</blockquote><blockquote>"vlcnfml1, three-quarter from mid-thigh up, straight-on, crouched stance, hands on thighs, stern gaze, red velvet curtain backdrop, studio lighting, even illumination, focused posture, dramatic contrast, professional demeanor, poised readiness."</blockquote><blockquote>"vlcnfml1, chest-up, eye-level, holding books, smiling, garden with roses, warm sunlight filtering through foliage, soft focus background, gentle breeze, peaceful moment, natural setting."</blockquote><blockquote>"vlcnfml1, full length from behind, low angle, walking away, hands relaxed at sides, city skyline reflected in water, bridge railing visible, warm streetlights and neon glow illuminate scene."</blockquote><h3>Training</h3><ul><li>Base model: Krea 2 (trained on Raw)</li><li>Type: Character LoRA</li><li>Dataset: 40 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: e5e1514e9a7e777c0bad3cea097a17f62a9f0ab04d5781d2c195e789c46855e4