Overview

This model represents Norwegian ethnicity, trained on an extensive set of captions describing young women identified as Norwegian. Compared to the Icelandic and Finnish datasets, this is the largest and most phenotype-diverse of the three, with a very strong Nordic core but a clearly learned range of secondary traits.

The model captures facial anatomy, skin tone, hair and eye color variation, and body build, and functions as a soft ethnicity anchor rather than a strict lock.

Facial Characteristics & Phenotype

Dominant and strongly reinforced traits:

• High or prominent cheekbones

• Symmetrical facial structure

• Narrow to oval face shapes

• Straight noses (small to medium, occasionally slightly upturned)

• Defined but not heavy jawlines

• Almond-shaped eyes (extremely common across all eye colors)

• Slightly pointed or softly rounded chins

Common variants:

• Rounder faces with prominent cheekbones

• Slightly angular or rectangular face shapes

• Softer jawlines with fuller cheeks

The almond-shaped eye contour + cheekbone prominence is one of the strongest and most consistent phenotype signals in the entire dataset.

Skin Tone & Undertones

This dataset contains the widest skin-tone range of the Nordic sets.

Dominant range:

• Very fair skin

• Fair skin

• Light skin tones

• Subtle pink undertones

• Neutral undertones

• Light warm undertones

Clearly present and learnable:

• Light olive skin

• Light brown skin

• Light-to-medium skin with olive undertones

These darker or olive tones appear frequently enough that the model will reproduce them unless constrained by prompt.

Hair Color & Texture

Hair color coverage is broad and heavily reinforced:

Very common:

• Blonde (platinum, light, golden, sandy)

• Light brown

• Medium brown

Common:

• Dark brown

• Red

• Auburn

• Strawberry blonde

Present but less dominant:

• Very dark brown (near-black)

Hair texture includes:

• Straight (most common)

• Wavy (very common)

• Curly (present, especially with brown or red hair)

Short hair, bangs, ponytails, and pulled-back styles are frequently represented, making this model flexible for modern looks.

Eye Color Range

Eye color is diverse and well learned:

Very common:

• Blue

• Light blue

• Blue-gray

Common:

• Green

• Blue-green

• Light brown

Present and consistent:

• Brown

• Dark brown

Eye color does not strongly correlate with hair color in this dataset, which explains why unconstrained prompts can drift.

Freckles & Skin Details

Freckles are frequently and explicitly described, especially:

• Across cheeks and nose

• On forehead and upper lip

• Occasionally across shoulders and arms

This makes freckles a reliable optional feature when prompting.

Body Build & Proportions

Strongly reinforced body traits:

• Slender build

• Slim or lean physique

• Narrow waist

• Visible collarbones

• Light athletic definition

Also present:

• Slight hourglass figures

• Petite builds

• Lean athletic builds

Heavy, stocky, or bulky body types are not represented.

Recommended Use

This model works best for:

• Portraits and headshots

• Upper-body or full-body figures

• Neutral or daylight lighting

• Explicit hair and eye color prompting

Because of the wide phenotype spread, explicit constraints are recommended if you want stereotypically Nordic outputs.