Sign In

Feral_Victoria

68

Loaded

Download

1 variant available

bf16 SafeTensor

Feral_Negative_Victoria_Krea2.safetensors

BF16, good balance • 218 MB

Verified:

Type
LoRA
Stats

154

68

336

Generation

Loaded

Reviews
Published

Oct 5, 2026

Base Model

Krea 2

Training
Steps: 2,500
Epochs: 15
Usage Tips
Strength: 1
Hash
AutoV2
CC7FE19424
Trigger Words
Victoria
Victoria, adult woman, short espresso-brown bob, amber-brown hooded eyes, medium-light olive skin, tall slim athletic build, small natural breasts
default creator card background decoration
Followers - 141

141

Likes - 205

205

Victoria 01.jpg

Victoria is a photorealistic character LoRA designed around a consistent adult female character with a short espresso-brown bob, amber-brown eyes, medium-light olive skin, refined angular features, and a tall, slim athletic figure.

Her appearance is designed to remain recognizable across a wide variety of environments, outfits, hairstyles, poses, expressions, and lighting conditions without forcing every generation toward the same composition or aesthetic.

Character Appearance

Victoria's core traits include:

  • short espresso-brown bob

  • amber-brown eyes

  • medium-light olive skin

  • mature adult facial features

  • naturally full lips

  • tall, slim athletic build

  • long legs

  • subtle waist

  • lean hips and thighs

  • small natural breasts

  • balanced, elegant proportions

Her overall look is mature, sophisticated, confident, and naturally sensual

Trigger Word

Victoria

LoRA Weight

A good starting range is:

0.8–1.0

Recommended starting point:

1.0

Prompting Victoria — This Is the Important Stuff!

A simple identity description works well as a starting point:

Victoria, adult woman, short espresso-brown bob, amber-brown hooded eyes, medium-light olive skin, tall slim athletic build, small natural breasts

For generations where body proportions are important, you can expand this by adding:

long legs, subtle waist, lean hips and thighs, small natural breasts, elegant balanced proportions

Victoria works well across a broad range of styles, including:

  • clean editorial portraits

  • tailored trousers and minimalist fashion

  • modern architectural environments

  • refined evening styling

  • tasteful lingerie

  • intimate lifestyle photography

  • mature themed beauty portraits

While different hairstyles can work, her natural hairstyle is a blunt bob cut. Her hair can be, softly tousled, deep side parted, tucked-behind-the-ears, slicked-back, and subtle braids work well too.

Victoria tends to look best when prompts emphasize:

  • natural skin texture

  • realistic photography

  • mature feminine styling

  • believable anatomy

  • tall, lean proportions

If your base model tends to exaggerate proportions, particularly chest size or body curves, you may need to reinforce phrases such as:

small natural breasts, tall slim athletic build, lean hips and thighs, balanced natural proportions

Victoria is intentionally designed with a lean athletic silhouette rather than an exaggerated hourglass figure, so reinforcing those proportions can help when using base models with stronger body-shape biases.

Base Model / Face Detailer Tip

Some community Krea2 base models can have a very strong built-in aesthetic bias.

This is especially noticeable with models trained heavily around adult-content or models with strong built-in demographic priors. In those cases, the base model may partially override Victoria's face even when the LoRA itself is working correctly.

If the overall image, pose, and body look good but the face has drifted away from Victoria, try running a second pass or Face Detailer pass with Victoria's LoRA applied again. This can help pull the facial structure, eye shape, coloring, and overall identity back toward Victoria while preserving the rest of the generation.

This is most useful for:

  • full-body or distant shots

  • adult-oriented base models

  • heavily stylized community models

  • models with a strong built-in facial aesthetic

  • images where her face occupies only a small part of the frame

LoRA strength during the face pass may need to be adjusted depending on the base model and workflow.