README.md

---
tags:
- flux.2
- image-to-image
- virtual-try-off
- fal
- lora
- diffusers
- template:diffusion-lora
widget:
- output:
    url: images/5.png
  text: >-
    TRYOFF extract the t-shirt over a white background, product photography style. 
    NO HUMAN VISIBLE (the garments maintain their 3D form like an invisible mannequin).
- output:
    url: images/3.png
  text: >-
    TRYOFF extract the dress over a white background, product photography style. 
    NO HUMAN VISIBLE (the garments maintain their 3D form like an invisible mannequin).
- output:
    url: images/2.png
  text: >-
    TRYOFF extract the pants over a white background, product photography style. 
    NO HUMAN VISIBLE (the garments maintain their 3D form like an invisible mannequin).
- output:
    url: images/6.png
  text: >-
    TRYOFF extract the outfit over a white background, product photography style. 
    NO HUMAN VISIBLE (the garments maintain their 3D form like an invisible mannequin).
- output:
    url: images/1.png
  text: >-
    TRYOFF extract the t-shirt over a white background, product photography style. 
    NO HUMAN VISIBLE (the garments maintain their 3D form like an invisible mannequin).
# - output:
#     url: images/4.png
#   text: >-
#     TRYOFF extract the upper body over a white background, product photography style. 
#     NO HUMAN VISIBLE (the garments maintain their 3D form like an invisible mannequin).
- output:
    url: images/7.png
  text: >-
    TRYOFF extract full outfit in the reference image over a white background,
    high-end professional product photography. Present the outfit as a complete,
    vertically stacked ensemble arranged as if worn. The items are stacked as if
    worn. The top-layer garment is dominant, followed directly by the
    bottom-layer garment. The footwear is placed below the bottom-layer hem,
    aligning with where the feet would naturally be. Lighting: Clean, even,
    diffused studio lighting (softbox or beauty dish style). The illumination
    must highlight all varying textures (e.g., pebble leather, suede, knit, or
    canvas) without creating harsh shadows.
base_model: black-forest-labs/FLUX.2-klein-9B
instance_prompt: TRYOFF
license: apache-2.0
pipeline_tag: image-to-image
---
# FLUX.2-klein-base-9B Virtual Try-Off LoRA

<Gallery />

Virtual Try-Off: Given an image of a person wearing clothing and a garment category prompt, the model generates a clean image of the garment as if it were photographed alone.\
The model reconstructs the clothing item while preserving its style, texture, color, and design from the input image.\
1 input image (person wearing clothes) + text category → 1 output garment image\
Built with fal.ai.

### Usage
- Try the LoRA on [fal.ai Playground](https://fal.ai/models/fal-ai/flux-2-klein/9b/base/edit/lora/playground?share=033cae58-12b5-4fef-992b-d48ff6759e01)
- Try the workflow (image-to-video) at [fal.ai Workflow](https://fal.ai/workflows/rizavelioglu/workflow-virtual-try-off?view=playground)
- ComfyUI: download compatible weights [virtual-tryoff-lora_comfy.safetensors](https://huggingface.co/fal/virtual-tryoff-lora/resolve/main/virtual-tryoff-lora_comfy.safetensors)
- Diffusers:
  ```python
  import torch
  from diffusers import Flux2KleinPipeline
  from PIL import Image
  
  pipeline = Flux2KleinPipeline.from_pretrained(
    "black-forest-labs/FLUX.2-klein-base-9B", 
    torch_dtype=torch.bfloat16, 
    low_cpu_mem_usage=False
  ).to("cuda")
  pipeline.load_lora_weights(
      "fal/virtual-tryoff-lora", 
      weight_name="virtual-tryoff-lora_diffusers.safetensors", 
      adapter_name="vtoff"
  )
  pipeline.set_adapters("vtoff", adapter_weights=1.0)
  pipeline.fuse_lora(adapter_names=["vtoff"], lora_scale=1.0)
  
  image = pipeline(
      image=Image.open("<your_image>.jpg"),
      prompt="TRYOFF extract the full outfit over a white background, product photography style.  NO HUMAN VISIBLE (the garments maintain their 3D form like an invisible mannequin).",
      height=1024,
      width=768,
      num_inference_steps=28,
      guidance_scale=5.0,
      generator=torch.Generator("cuda").manual_seed(42),
  ).images[0]
  ```

### Training
Trained with [fal.ai trainer](https://fal.ai/models/fal-ai/flux-2-klein-9b-base-trainer/edit).
- Base model: FLUX.2-klein-base-9B
- Steps: 10000
- Learning Rate: 0.00005
- Dataset: 300 image pairs (model + garment) of shape `1024x1024`

### Author
Created by [Riza Velioglu](https://rizavelioglu.github.io/) at [fal.ai](https://fal.ai/)