README.md

---
tags:
- text-to-image
- lora
- diffusers
- template:diffusion-lora
widget:
- text: >-
    A kitsune woman, fox ears, fox tail, anime, female, short hair, black and
    blue hair, red highlights, bold makeup, red lips, piercing gaze, close-up
    portrait, detailed accessories, choker necklace, layered necklaces,
    earrings, gothic fashion, dramatic lighting, dark background, expressive
    eyes, soft skin, emotional expression, beautiful face, stylish, glamorous,
    vivid colorsBlonde woman with long wavy hair, sitting in a lush forest, soft
    golden sunlight filtering through trees, wearing a flowing white robe with
    intricate black embroidery, eyes closed in serenity, peaceful expression,
    seated on grass with dappled sunlight, tall trees surrounding her, gentle
    breeze slightly rustling her hair, early morning or late afternoon light,
    tranquil and mystical atmosphere, slightly out-of-focus background enhancing
    subject focus, warm color palette, soft shadows and highlights creating
    depth.  <lora:Kitsune_v51:1>
  parameters:
    negative_prompt: 'Guidance: 1 Steps: 20 Seed: 770198788'
  output:
    url: images/00900-770198788.png
base_model: black-forest-labs/FLUX.1-dev
instance_prompt: kitsune
license: unknown
---
# Kitsune &#x2F; Fox Girl

<Gallery />

## Model description 

FROM https:&#x2F;&#x2F;civitai.com&#x2F;models&#x2F;1097605&#x2F;kitsune-fox-girl?modelVersionId&#x3D;1232918

Trigger kitsune
Strength 1

I was working on a commission and realized that Flux does not have a good kitsune concept, so I trained this LoRA to improve Flux&#39;s pre-trained concept. It fits my needs, though it still has to fight with Flux to correct the tail anatomy and animal ears (without adding additional human ears). It gets the concept and placement right about 50% of the time, which isn&#39;t bad. It will make a kitsune or fox girl in several styles, including anime and semi-realistic. Kitsune have up to nine tails, so you may get a random number if you do not specify (and sometimes even if you do).

If anyone is interested, I&#39;ll refine this concept further. If you want to see similar concepts, post about your interest in the discussion.

Usage
To use the most recent version of the LoRA, use the following settings:

Trigger word: just kitsune will work, even if you are going for a fox girl

Other tokens that work well include mentioning her fox ears and fluffy fox tail, which will also help; you can change the color and, in some cases, the number of tails. However, not enough multi-tailed kitsune images were used in the training, so it may not work as well as intended.

Lora Strength: Use the LoRA at full strength.

## Trigger words

You should use `kitsune` to trigger the image generation.


## Download model

Weights for this model are available in Safetensors format.

[Download](/Jonjew/FoxGirl/tree/main) them in the Files & versions tab.