<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>Cardassian 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>crdssnfml1</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>"crdssnfml1, chest-up, high angle, selfie pose, slight smile, cream blouse, dark pants, stone alleyway, bright sunlight casting shadows on cobblestones and walls."</blockquote><blockquote>"crdssnfml1, three-quarter from mid-thigh up, straight-on, leaning on railing, neutral expression, beige linen outfit, theater marquee backdrop, night street lighting, colorful string lights, urban sidewalk, neon glow."</blockquote><blockquote>"crdssnfml1, chest-up, eye-level, standing, looking left, neutral expression, plaid shirt, garage with car and tools, soft natural light from side, shallow depth of field, realistic texture, focused on subject."</blockquote><blockquote>"crdssnfml1, full length, eye-level, walking forward, neutral expression, dark zip-up jacket, olive pants, bicycle workshop, natural light from window, tools and bikes surrounding, concrete floor."</blockquote><blockquote>"crdssnfml1, chest-up, straight-on, head bowed, gaze downward, solemn, dark void backdrop, soft directional lighting from front-left, subtle shadows accentuate contours, poised stillness, dramatic contrast."</blockquote><blockquote>"crdssnfml1, chest-up, eye-level, standing centered, smiling slightly, lush tropical foliage behind, soft golden sunlight filtering through leaves, warm glow highlighting figure against greenery."</blockquote><blockquote>"crdssnfml1, chest-up, straight-on, smiling broadly, head slightly tilted, olive backdrop, soft even lighting, confident stance, hands at sides, neutral expression, focused on upper body."</blockquote><blockquote>"crdssnfml1, chest-up, eye-level, hand on strap, head tilted down, serene expression, flower market backdrop, soft diffused light, vibrant blooms, wooden stalls, people browsing, overcast sky, natural ambient glow."</blockquote><blockquote>"crdssnfml1, chest-up, eye-level, turned slightly left, gazing forward, neutral expression, workshop with fabric rolls and mannequin, soft natural light from window, wooden shelves, sewing tools."</blockquote><blockquote>"crdssnfml1, three-quarter from mid-thigh up, eye-level, standing with one foot on truck step, mouth open, olive shirt, beige pants, brown boots, food truck park, bright daylight, casual outdoor setting, vibrant colorful trucks, people in background."</blockquote><blockquote>"crdssnfml1, three-quarter from mid-thigh up, low angle, seated cross-legged, hands resting on knees, neutral expression, dimly lit auditorium with star maps projected on ceiling, warm ambient lighting from wall sconces."</blockquote><blockquote>"crdssnfml1, close-up portrait, eye-level, facing forward, neutral expression, stone castle backdrop, warm golden hour light, soft focus background, calm demeanor, direct gaze, detailed textures visible."</blockquote><blockquote>"crdssnfml1, chest-up, eye-level, standing by riverbank, gazing forward, stern expression, rocky shore and flowing water behind, warm golden hour light reflecting off water surface."</blockquote><blockquote>"crdssnfml1, chest-up, eye-level, drinking from bottle, neutral expression, urban night bus stop, streetlights and building windows, warm overhead lamp casting soft glow on subject."</blockquote><blockquote>"crdssnfml1, chest-up, eye-level, laughing broadly, wearing knit cardigan and dark top, standing by river at night, city lights reflected, warm ambient glow from buildings and street lamps."</blockquote><blockquote>"crdssnfml1, three-quarter from mid-thigh up, eye-level, walking forward with arms bent, smiling, floral dress, rolling hills with dirt path, soft golden hour light."</blockquote><blockquote>"crdssnfml1, chest-up, slightly low angle, holding apple near mouth, neutral expression, bedroom with bed, soft ambient light, quiet moment, indoor setting, subtle shadows, relaxed posture, focused gaze."</blockquote><blockquote>"crdssnfml1, three-quarter from mid-thigh up, side angle, crouching, mouth open, intense expression, beach with ocean, bright daylight, sand scattered, horizon visible, strong shadows cast, dynamic posture."</blockquote><blockquote>"crdssnfml1, three-quarter from mid-thigh up, eye-level, leaning against rock, smiling broadly, waterfall cascading behind, natural daylight, rugged stone textures, water mist."</blockquote><blockquote>"crdssnfml1, three-quarter from mid-thigh up, slightly low angle, seated with hands clasped, smiling gently, surrounded by clocks and tools, warm lamp glow illuminates workspace, wooden walls, soft directional lighting."</blockquote><h3>Training</h3><ul><li>Base model: Krea 2 (trained on Raw)</li><li>Type: Character LoRA</li><li>Dataset: 28 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: 29e1360f3e43d47a39035a409a4174ea30dd91fbb0dad3931fac4f20998f5101