---
license: other
license_name: bria-legal-lobby
license_link: https://bria.ai/legal-lobby
---


# BRIA-3.2 ControlNet Union Model Card

BRIA-3.2 ControlNet-Union, trained on the foundation of [BRIA-3.2 Text-to-Image](https://huggingface.co/briaai/BRIA-3.2), supports 6 control modes, including depth (0), canny (1), colorgrid (2), recolor (3), tile (4), pose (5). This model can be jointly used with other ControlNets.

Built with a strong commitment to legal compliance and responsible AI practices, this model ensures safe and scalable generative image capabilities for commercial use.


[CLICK HERE FOR A DEMO](https://huggingface.co/spaces/briaai/BRIA-3.2-ControlNet-Union)

For more information, please visit our [website](https://bria.ai/).

Join our [Discord community](https://discord.gg/Nxe9YW9zHS) for more information, tutorials, tools, and to connect with other users!

### Get Access
BRIA-3.2-ControlNet-Union requires access to BRIA-3.2 Text-to-Image. For more information, [click here](https://huggingface.co/briaai/BRIA-3.2).
  
### Model Description
- **Developed by:** BRIA AI
- **Model type:** Latent Flow-Matching Text-to-Image Model
- **License:** [Commercial licensing terms & conditions.](https://bria.ai/customer-general-terms-and-conditions)
- Purchase is required to license and access the model.
- **Model Description:** ControlNet Union for BRIA-3.2 Text-to-Image model. The model generates images guided by text and a conditioned image.
- **Resources for more information:** [BRIA AI](https://bria.ai/)


## Control Mode
| Control Mode | Description |
|:------------:|:-----------:|
|0|depth
|1|canny
|2|colorgrid
|3|recolor
|4|tlie
|5|pose


```python

```

### Installations
```bash
pip install -qr https://huggingface.co/briaai/BRIA-3.2/resolve/main/requirements.txt
pip install diffusers==0.30.2, hf_hub_download
```

```python
from huggingface_hub import hf_hub_download
import os
try:
    local_dir = os.path.dirname(__file__)
except:
    local_dir = '.'
    
hf_hub_download(repo_id="briaai/BRIA-3.2", filename='pipeline_bria.py', local_dir=local_dir)
hf_hub_download(repo_id="briaai/BRIA-3.2", filename='transformer_bria.py', local_dir=local_dir)
hf_hub_download(repo_id="briaai/BRIA-3.2", filename='bria_utils.py', local_dir=local_dir)
hf_hub_download(repo_id="briaai/BRIA-3.2-ControlNet-Union", filename='pipeline_bria_controlnet.py', local_dir=local_dir)
hf_hub_download(repo_id="briaai/BRIA-3.2-ControlNet-Union", filename='controlnet_bria.py', local_dir=local_dir)
```

# Inference
```python
import torch
from diffusers.utils import load_image
from controlnet_bria import BriaControlNetModel
from pipeline_bria_controlnet import BriaControlNetPipeline
import PIL.Image as Image

RATIO_CONFIGS_1024 = {
    0.6666666666666666: {"width": 832, "height": 1248},
    0.7432432432432432: {"width": 880, "height": 1184},
    0.8028169014084507: {"width": 912, "height": 1136},
    1.0: {"width": 1024, "height": 1024},
    1.2456140350877194: {"width": 1136, "height": 912},
    1.3454545454545455: {"width": 1184, "height": 880},
    1.4339622641509433: {"width": 1216, "height": 848},
    1.5: {"width": 1248, "height": 832},
    1.5490196078431373: {"width": 1264, "height": 816},
    1.62: {"width": 1296, "height": 800},
    1.7708333333333333: {"width": 1360, "height": 768},
}

def resize_img(control_image):
    image_ratio = control_image.width / control_image.height
    ratio = min(RATIO_CONFIGS_1024.keys(), key=lambda k: abs(k - image_ratio))
    to_height = RATIO_CONFIGS_1024[ratio]["height"]
    to_width = RATIO_CONFIGS_1024[ratio]["width"]
    resized_image = control_image.resize((to_width, to_height), resample=Image.Resampling.LANCZOS)
    return resized_image


base_model = 'briaai/BRIA-3.2'
controlnet_model = 'briaai/BRIA-3.2-ControlNet-Union'
controlnet = BriaControlNetModel.from_pretrained(controlnet_model, torch_dtype=torch.bfloat16)
pipeline = BriaControlNetPipeline.from_pretrained(base_model, controlnet=controlnet, trust_remote_code=True)
pipeline = pipeline.to(device="cuda", dtype=torch.bfloat16)

control_image_canny = load_image("https://huggingface.co/briaai/BRIA-3.2-ControlNet-Union/resolve/main/images/canny.jpg")
controlnet_conditioning_scale = 1.0
control_mode = 1
control_image_canny = resize_img(control_image_canny)
width, height = control_image_canny.size

prompt = 'In a serene living room, someone rests on a sapphire blue couch, diligently drawing in a rose-tinted notebook, with a sleek black coffee table, a muted green wall, an elegant geometric lamp, and a lush potted palm enhancing the peaceful ambiance.'

generator = torch.Generator(device="cuda").manual_seed(555)
image = pipeline(
    prompt, 
    control_image=control_image_canny,
    control_mode=control_mode,
    width=width,
    height=height,
    controlnet_conditioning_scale=controlnet_conditioning_scale,
    num_inference_steps=50, 
    max_sequence_length=128,
    guidance_scale=5,
    generator=generator,
    negative_prompt="Ugly,Morbid,Extra fingers,Poorly drawn hands,Mutation,Blurry,Extra limbs,Gross proportions,Missing arms,Mutated hands,Long neck,Duplicate"
).images[0]
print(image)        
```

# Multi-Controls Inference
```python
import torch
from diffusers.utils import load_image
from controlnet_bria import BriaControlNetModel, BriaMultiControlNetModel
from pipeline_bria_controlnet import BriaControlNetPipeline
import PIL.Image as Image

base_model = 'briaai/BRIA-3.2'
controlnet_model = 'briaai/BRIA-3.2-ControlNet-Union'

controlnet = BriaControlNetModel.from_pretrained(controlnet_model, torch_dtype=torch.bfloat16)
controlnet = BriaMultiControlNetModel([controlnet])

pipe = BriaControlNetPipeline.from_pretrained(base_model, controlnet=controlnet, torch_dtype=torch.bfloat16, trust_remote_code=True)
pipe.to("cuda")

control_image_colorgrid = load_image("https://huggingface.co/briaai/BRIA-3.2-ControlNet-Union/resolve/main/images/colorgrid.jpg")
control_image_pose = load_image("https://huggingface.co/briaai/BRIA-3.2-ControlNet-Union/resolve/main/images/pose.jpg")

control_image = [control_image_colorgrid, control_image_pose]
controlnet_conditioning_scale = [0.5, 0.5]
control_mode = [2, 5]

width, height = control_image[0].size

prompt = 'Two kids in jackets play near a tent in a forest.'

generator = torch.Generator(device="cuda").manual_seed(555)
image = pipe(
    prompt, 
    control_image=control_image,
    control_mode=control_mode,
    width=width,
    height=height,
    controlnet_conditioning_scale=controlnet_conditioning_scale,
    num_inference_steps=50, 
    max_sequence_length=128,
    guidance_scale=5,
    generator=generator,
    negative_prompt="Ugly,Morbid,Extra fingers,Poorly drawn hands,Mutation,Blurry,Extra limbs,Gross proportions,Missing arms,Mutated hands,Long neck,Duplicate"
).images[0]
```