---
license: other
license_name: stabilityai-ai-community
license_link: LICENSE.md
language:
- en
base_model:
- stabilityai/stable-diffusion-3.5-medium
pipeline_tag: text-to-image
---

<div align="center">

**TensorArt Stable Diffusion 3.5 Medium ControlNet Canny**

<img src="./asset/demo.png"/>
</div>

# With SD3.5M
```python
import torch
from diffusers import StableDiffusion3ControlNetPipeline
from diffusers import SD3ControlNetModel
from diffusers.utils import load_image

controlnet = SD3ControlNetModel.from_pretrained("tensorart/SD3.5M-Controlnet-Canny")
pipe = StableDiffusion3ControlNetPipeline.from_pretrained(
    "stabilityai/stable-diffusion-3.5-medium",
    controlnet=controlnet
)
pipe.to("cuda", torch.float16)

control_image = load_image("https://huggingface.co/tensorart/SD3.5M-Controlnet-Canny/resolve/main/asset/canny.png")
prompt = "An eye"
negative_prompt = "low quality, worst quality, deformed, distorted, disfigured, motion smear, motion artifacts, fused fingers, bad anatomy, weird hand, ugly, monochrome"

image = pipe(
    prompt, 
    num_inference_steps=30,
    negative_prompt=negative_prompt, 
    control_image=control_image, 
    guidance_scale=4.5,
    controlnet_conditioning_scale=0.8
).images[0]
image.save('image.jpg')
```


# With TensorArt's SD3.5M Turbo
```python
import torch
from diffusers import StableDiffusion3ControlNetPipeline
from diffusers import SD3ControlNetModel
from diffusers.utils import load_image

controlnet = SD3ControlNetModel.from_pretrained("tensorart/SD3.5M-Controlnet-Canny")
pipe = StableDiffusion3ControlNetPipeline.from_pretrained(
    "tensorart/stable-diffusion-3.5-medium-turbo",
    controlnet=controlnet
)
pipe.to("cuda", torch.float16)

control_image = load_image("https://huggingface.co/tensorart/SD3.5M-Controlnet-Canny/resolve/main/asset/canny.png")
prompt = "An eye"
negative_prompt = "low quality, worst quality, deformed, distorted, disfigured, motion smear, motion artifacts, fused fingers, bad anatomy, weird hand, ugly, monochrome"

image = pipe(
    prompt, 
    num_inference_steps=8,
    negative_prompt=negative_prompt, 
    control_image=control_image, 
    guidance_scale=1.5,
    controlnet_conditioning_scale=0.8
).images[0]
image.save('image.jpg')
```

## Contact
* Website: https://tensor.art  https://tusiart.com
* Developed by: TensorArt