---
license: other
license_name: bespoke-lora-trained-license
license_link: https://multimodal.art/civitai-licenses?allowNoCredit=True&allowCommercialUse=RentCivit&allowDerivatives=False&allowDifferentLicense=True
tags:
- text-to-image
- stable-diffusion
- lora
- diffusers
- template:sd-lora
- migrated
- photorealistic
- woman
- actress
- celebrity
- girls
- realistic
- celebrity,

base_model: black-forest-labs/FLUX.1-dev
instance_prompt: ohwx
widget:
- text: 'Breathtaking over the shoulder shot photography of ohwx looking at viewerOff-shoulder silky blouse with ruched bust and flared sleeves , necklace, eyelashes, fine hair detail, entire hairstyle visible, perfect eyes with iris pattern, sensual lips, nose, (perfectly sharp:1.3), realistic textures, (deep focus on subject, blurred background:1.4), 8k uhd, dslr, ultra high quality image, film grain, Fujifilm XT3'
  parameters:
    negative_prompt: KatherineHeigl_flux_lora_v1_Weight-1.0
            
  output:
    url: >-
      49761287.jpeg
- text: 'Breathtaking medium shot photography of ohwx, (Corset top with boning and satin ribbon lace-up:1.3) , Townhouse,style by René Laloux,soft lighting, upper body shot, collarbone, smile, (upper body framing:1.3), bare shoulder, detailed texture quality wool top, simple black choker, sensual lips, eyelashes, (entirely visible beautiful hairstyle:1.2), fine hair detail, perfect eyes, iris pattern, eyes makeup, (perfectly sharp:1.3), (head to shoulders composition:1.4), background is a blurry flower field, realistic textures, (deep focus:1.1), negative space around subject, 8k uhd, dslr, ultra high quality image, film grain, Fujifilm XT3'
  parameters:
    negative_prompt: KatherineHeigl_flux_lora_v1_Weight-1.1
            
  output:
    url: >-
      49761301.jpeg
- text: 'Breathtaking over the shoulder shot photography of ohwx looking at viewer, high collar white blouse, imperfections, necklace with ornament falling down her back, looking over shoulders, eyelashes, fine hair detail, entire hairstyle visible, perfect eyes with iris pattern, sensual lips, nose, (perfectly sharp:1.3), realistic textures, (deep focus on subject, blurred background:1.4), 8k uhd, dslr, ultra high quality image, film grain, Fujifilm XT3'
  parameters:
    negative_prompt: KatherineHeigl_flux_lora_v1_Weight-1.0
            
  output:
    url: >-
      49761313.jpeg

---

# 👑 Katherine Heigl (90s)(Flux) 🎬 

<Gallery />



([CivitAI](https://civitai.com/models/))

## Model description

<h1 id="***-if-you-love-it-like-it!-***-vva150sqr">👍 <span style="color:rgb(64, 192, 87)">***</span> <strong><em><span style="color:rgb(34, 139, 230)">If you love it, like it!</span></em></strong> <span style="color:rgb(64, 192, 87)">***</span>👍</h1><p><em>workflow: </em><a target="_blank" rel="ugc" href="https://civitai.com/models/1088678"><em>https://civitai.com/models/1088678</em></a></p><h1 id="katherine-heigl-e23hl0afj">👑 <span style="color:rgb(121, 80, 242)">Katherine Heigl </span>🎬</h1><p><strong>About my celebrities loras</strong><br />90% of the dataset used to build my loras only use head images. That really help the blend with other lora or model as there is no hands, feet, that may or will interfere in the final image render. When you get distorted hands with a person lora, it's because there is info on hands in the dataset used to train the lora, but that will not happen with my loras.</p><p>I've trained on Flux.1 Dev so other merged or trained checkpoint may not work well with my loras.</p><p>The drawback side of that is that the body may not be reflecting the reality. It may not be a drawback tho.</p><p></p><p>This is a lora for Flux.1 Dev. Work with other model but you must drop some simple bloc (good start 19-32).</p><p>Trained with ai-toolkit, so merging it is not easy.</p><p></p><p><strong>To get the best result</strong></p><p><em>Guidance: </em><strong><em>2.2-3</em></strong></p><p><em>Steps (dev): </em><strong><em>30-40</em></strong></p><p><em>daemon detailer (lying sigma sampler): factor: -0.02, start 0.06, end 0.75</em></p><p><em>Resolution: </em><strong><em>Upscale the latent by 1.25 or 1.5</em></strong><em> you'll get awsome result. (take longer time but worth it)</em></p><p><em>Trigger word is (may work better in certain context): </em><strong><em>ohwx</em></strong></p><p></p><p>Enjoy!</p>

## Trigger words
You should use `ohwx` to trigger the image generation.
    

## Download model

Weights for this model are available in Safetensors format.

[Download](/Keltezaa/katherine-heigl-90s-flux/tree/main) them in the Files & versions tab.

## Use it with the [🧨 diffusers library](https://github.com/huggingface/diffusers)

```py
from diffusers import AutoPipelineForText2Image
import torch

device = "cuda" if torch.cuda.is_available() else "cpu"

pipeline = AutoPipelineForText2Image.from_pretrained('black-forest-labs/FLUX.1-dev', torch_dtype=torch.bfloat16).to(device)
pipeline.load_lora_weights('Keltezaa/katherine-heigl-90s-flux', weight_name='KatherineHeigl_flux_lora_v1.safetensors')
image = pipeline('Breathtaking over the shoulder shot photography of ohwx looking at viewer, high collar white blouse, imperfections, necklace with ornament falling down her back, looking over shoulders, eyelashes, fine hair detail, entire hairstyle visible, perfect eyes with iris pattern, sensual lips, nose, (perfectly sharp:1.3), realistic textures, (deep focus on subject, blurred background:1.4), 8k uhd, dslr, ultra high quality image, film grain, Fujifilm XT3').images[0]
```

For more details, including weighting, merging and fusing LoRAs, check the [documentation on loading LoRAs in diffusers](https://huggingface.co/docs/diffusers/main/en/using-diffusers/loading_adapters)