A LoRA for generating Norwegianwomen across a range of hair colors, eye colors, and hairstyles. It works well with simple photographic prompts and clear descriptions of the features you want.

Trigger word

No trigger word was used in the example tests.

How to use

Start at strength 0.8 for portraits or 1.0 for a stronger effect. Both strengths were tested successfully, although they were not compared with matching prompts across every test.

Describe the subject, hair color, eye color, expression, and framing in ordinary language. Use one clear term for each feature.

Prompt syntax

Photographic [headshot / portrait / full-body photograph] of a young woman with [hair color] hair, [eye color] eyes, and [hairstyle]. [Expression or gaze]. [Background]. [Framing].

Example prompts

Front-facing photographic headshot of a young woman with chestnut brown hair and green eyes. Straight shoulder-length hair with a center part. Neutral expression, plain white background, full head and hairline visible.

Photographic portrait of a young woman with copper red hair in a chin-length bob, blue eyes, and a slight smile. Plain gray background, shoulders-up framing.

Full-body studio photograph of a young woman wearing a black two-piece bikini. Standing naturally with arms relaxed against a plain light gray background. Wide head-to-toe composition with her entire head and both feet visible.

Features that responded well

Hair colors showed clear variation between black, brown, blonde, and red. Platinum blonde, golden blonde, and copper red were particularly easy to distinguish.

Bobs, bangs, center parts, ponytails, and half-up styles generally came through clearly. Straight hair, loose waves, and crimping were also visibly different.

Blue and dark brown were the easiest eye colors to distinguish. Close-up eye prompts made the differences easier to inspect.

Limitations

Black hair combined with dark brown eyes sometimes changed the facial appearance substantially, even with the same seed. Expect possible changes to facial features as well as the requested colors. These renders do not establish ethnicity.

Similar shades can overlap. Jet black and soft black often looked alike, as did neighboring brown shades. Gray, green, olive green, hazel, and amber eyes sometimes produced similar results.

Ringlets, corkscrew curls, and coils often converged on similar curls. Deep-wave prompts sometimes produced curls instead.

Silver-gray hair made subjects appear older even when the prompt said young woman. Full-body prompts sometimes cropped out the feet or moved to a closer portrait. Exact iris color is harder to judge in full-body images.

Skin-tone wording changed the visible complexion, but neighboring beige shades were subtle. Rosy wording often produced strong cheek redness.

Tips

For eye-color tests, use a front-facing headshot with both irises clearly visible.

For hairstyle tests, leave space above the crown and beside the head.

For full-body images, use a wider camera distance and explicitly include both feet.

Use a fixed seed when comparing wording, then try additional seeds to check consistency.

About the examples

The tests used square headshots, portrait headshots, eye close-ups, and full-body images. The observations describe visible prompt adherence in those renders. Ancestry cannot be verified from appearance, and these tests did not include a LoRA-off control.