<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>Tellarite 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>tllrtfml1</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>"tllrtfml1, chest-up, eye-level, leaning on railing, gazing forward, neutral expression, beach at dusk, campfire glow and string lights, ambient evening light, wooden railing, sand, ocean waves, distant trees."</blockquote><blockquote>"tllrtfml1, chest-up, slightly low angle, leaning forward with hands on bag, intense gaze, dark coat and scarf, bustling night market stalls, warm artificial lights from overhead and signs."</blockquote><blockquote>"tllrtfml1, waist-up, eye-level, hands clasped mid-stride, neutral expression, dark turtleneck, dark pants, orchard rows, golden hour backlighting, soft focus foliage, warm ambient glow, natural depth."</blockquote><blockquote>"tllrtfml1, chest-up, eye-level, seated relaxed, slight smile, varsity jacket, comic book shelves, window light, casual pose, comic-themed backdrop, natural illumination, indoor setting."</blockquote><blockquote>"tllrtfml1, full length, straight-on, hands on hips, neutral expression, wine barrels in rows, bright overhead lighting, concrete floor, industrial setting, vertical light source, clean lines, uniform background."</blockquote><blockquote>"tllrtfml1, full length from behind, straight-on, standing still, facing neon sign, urban night scene, wet pavement, vibrant glowing signage, ambient glow reflecting on ground, warm interior lights visible through windows."</blockquote><blockquote>"tllrtfml1, full length, eye level, holding camera at chest height, focused gaze, bamboo forest backdrop, dappled sunlight filtering through tall stalks, natural shadows, quiet concentration, standing amidst tall greenery."</blockquote><blockquote>"tllrtfml1, chest-up, straight-on, facing forward, neutral expression, dark backdrop, soft directional lighting, subtle shadows, focused on upper torso."</blockquote><blockquote>"tllrtfml1, chest-up, eye-level, standing, facing forward, stern gaze, coastal pier, ocean backdrop, warm golden hour light, soft shadows, serene yet commanding presence, wooden planks beneath feet, distant cliffs visible."</blockquote><blockquote>"tllrtfml1, chest-up, eye-level, standing, facing slightly right, neutral expression, rural barn background, overcast daylight, earthy ground, vintage tractor visible, red barn with white trim."</blockquote><blockquote>"tllrtfml1, three-quarter from mid-thigh up, slightly low angle, seated relaxed on patterned rug, smiling broadly, lantern-lit alleyway backdrop, warm ambient glow from hanging lights, soft golden illumination casting gentle shadows."</blockquote><blockquote>"tllrtfml1, extreme close-up, frontal, seated, mouth open, intense expression, theater seats background, bright overhead lighting, red upholstery, rows of empty seats, ceiling lights visible, focused on subject's upper body."</blockquote><blockquote>"tllrtfml1, chest-up, slightly low angle, head tilted back laughing, wearing striped sweater and jeans, inside wooden boat cabin with map and lamp, soft warm light from hanging lantern."</blockquote><blockquote>"tllrtfml1, close-up portrait, eye level, head slightly tilted, eyes closed, red plaid shirt, outdoor concrete area, overcast daylight, soft natural illumination, calm demeanor, neutral background, shallow depth of field."</blockquote><blockquote>"tllrtfml1, three-quarter from mid-thigh up, slightly low angle, standing by railing, hands in pockets, slight smile, river and distant lights, soft golden hour glow, calm water, stone balustrade."</blockquote><blockquote>"tllrtfml1, three-quarter from mid-thigh up, slightly low angle, seated with arms raised, mouth open, expressive gesture, wooden ship interior, warm glow from hanging lamp, ropes, maps, portholes, rich wood paneling."</blockquote><blockquote>"tllrtfml1, full length, eye level, walking barefoot on forest floor, smiling, tall trees surrounding, dappled sunlight filtering through canopy, serene woodland path, gentle movement."</blockquote><blockquote>"tllrtfml1, chest-up, eye-level, standing, gazing forward, stern, beach at dusk, soft ambient light, distant cliffs, ocean waves, dark sand, scattered stones, serene yet commanding presence."</blockquote><blockquote>"tllrtfml1, close-up portrait, eye-level, facing slightly right, neutral expression, kitchen background, warm overhead lighting, blurred chefs working, metallic pots hanging, wooden surfaces, ambient glow from ceiling fixtures."</blockquote><blockquote>"tllrtfml1, chest-up, straight-on, standing, hands clasped, neutral expression, solid teal backdrop, even studio lighting, formal posture, professional portrait style, focused on upper body."</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: 7fffe3b167d44e812e017f64e4a36b931c5eeb7a1c28c358af5dce60e3471b14