This model represents Finnish ethnicity, trained on a large, highly repetitive set of captions describing young women identified as Finnish. The model captures facial structure, skin tone, hair and eye color distributions, and body build tendencies observed in the dataset, rather than enforcing a single Nordic stereotype.Compared to the Icelandic set, this dataset contains broader phenotype variance, including both Northern European features and a smaller but non-trivial number of darker-haired and darker-skinned individuals.<hr /><h2 id="facial-characteristics-and-phenotype">Facial Characteristics &amp; Phenotype</h2>The most consistently reinforced facial traits are:<ul><li>High or prominent cheekbones</li><li>Narrow to oval face shapes</li><li>Slightly angular or pointed chins</li><li>Straight noses (occasionally slightly upturned)</li><li>Defined but not heavy jawlines</li><li>Almond-shaped eyes (very common across all eye colors)</li><li>Symmetrical facial proportions</li></ul>Less common but present:<ul><li>Rounder face shapes</li><li>Softer jawlines</li><li>Slightly broader foreheads</li></ul>Overall, the facial phenotype leans slender, youthful, and balanced, with almond-shaped eyes being one of the strongest recurring signals.<hr /><h2 id="skin-tone-and-undertones">Skin Tone &amp; Undertones</h2>This dataset shows much wider skin-tone coverage than the Icelandic set.<h3 id="dominant-range:">Dominant range:</h3><ul><li>Very fair skin</li><li>Fair skin</li><li>Light skin tones</li><li>Subtle pink undertones</li><li>Neutral undertones</li><li>Light warm undertones</li></ul><h3 id="present-but-less-frequent:">Present but less frequent:</h3><ul><li>Light olive skin tones</li><li>Light brown skin tones</li><li>Medium-light skin with olive undertones</li></ul>These darker or olive tones appear consistently enough to be learned, but not strongly enough to dominate unless explicitly prompted.<hr /><h2 id="hair-color-and-texture">Hair Color &amp; Texture</h2>Hair color variation is broad and well represented:<h3 id="very-common:">Very common:</h3><ul><li>Blonde (light, medium, golden)</li><li>Strawberry blonde</li><li>Light brown</li><li>Medium brown</li></ul><h3 id="common:">Common:</h3><ul><li>Dark brown</li><li>Red / auburn</li><li>Copper red</li></ul><h3 id="rare-but-present:">Rare but present:</h3><ul><li>Black hair</li></ul>Hair texture is predominantly:<ul><li>Straight</li><li>Smooth</li><li>Fine to medium thickness</li></ul>Wavy hair appears regularly; curly hair is rare but present, almost exclusively paired with darker hair and darker skin tones.<hr /><h2 id="eye-color-range">Eye Color Range</h2>Eye color coverage is broad and well reinforced:<h3 id="common:">Common:</h3><ul><li>Blue</li><li>Light blue</li><li>Green</li><li>Blue-green</li><li>Light brown</li></ul><h3 id="frequent:">Frequent:</h3><ul><li>Brown</li><li>Dark brown</li></ul>Almond-shaped eyes appear across all eye colors, making eye shape a stronger learned signal than eye color itself.<hr /><h2 id="body-build-and-proportions">Body Build &amp; Proportions</h2>The captions repeatedly reinforce:<ul><li>Slender build</li><li>Slim or lean physique</li><li>Narrow waist</li><li>Long limbs</li><li>Light athletic definition (especially shoulders and arms)</li></ul>Some variation includes:<ul><li>Slight hourglass figures</li><li>Petite builds</li><li>Lean athletic builds</li></ul>Heavy or stocky body types are not represented.<hr /><h2 id="recommended-use">Recommended Use</h2>This model works best for:<ul><li>Portraits and headshots</li><li>Full-body or mid-body figures with neutral poses</li><li>Neutral or daylight lighting</li><li>Explicit hair and eye color prompting</li></ul>It functions as a soft ethnicity anchor, not a hard lock. Because the dataset includes real variation, ambiguous prompts may drift toward darker hair or eyes unless constrained.

Files with SHA256: 1372531b3ebada1111f6909a44096340a766286c3342333a38b9b30663b30728