---
tags:
- SDXL
- controlnet
- diffusers
- shadow generation
widget:
  - text: "a product with soft natural shadow, white background"
    output:
      url: pictures/pic1.png
  - text: "a product with hard natural shadow, white background"
    output:
      url: pictures/pic2.png
  - text: "a product with floating natural shadow, white background"
    output:
      url: pictures/pic3.png
  - text: "a product with soft natural shadow, white background"
    output:
      url: pictures/pic4.png
  - text: "a product with hard natural shadow, white background"
    output:
      url: pictures/pic5.png
  - text: "a product with floating natural shadow, white background"
    output:
      url: pictures/pic6.png
  - text: "a product with soft natural shadow, white background"
    output:
      url: pictures/pic7.png
  - text: "a product with hard natural shadow, white background"
    output:
      url: pictures/pic9.png
  - text: "a product with floating natural shadow, white background"
    output:
      url: pictures/pic8.png
base_model:
- stabilityai/stable-diffusion-xl-base-1.0
license: other
license_name: sdxl-non-commercial-license
---
<Gallery />

## Inference
```python
python3 generate_shadow_main.py
```


## Gradio Demo
```python
python3 app_shadow_generation_gradio.py
```


# Training

Training was done using https://github.com/huggingface/diffusers/blob/main/examples/controlnet/train_controlnet_sdxl.py