README.md

---
license: other
tags:
- controlnet
- text-to-image
- image-to-image
- inpainting
- qwen-image
- qwen-image-2512
- flux
- flux.1-dev
- quantized
- int8
- convrot
- comfyui
pipeline_tag: image-to-image
---

# ControlNet Models (ConvRot INT8)

High-fidelity **ConvRot INT8** quantized weights for ControlNet models across diverse architectures (Qwen-Image, FLUX.1-dev).

---

## 🌟 Model Overview

This repository hosts high-quality **ConvRot INT8** quantized weights for multi-condition ControlNet Union models, designed to significantly reduce VRAM and storage footprint while preserving fine-grained structural control fidelity:

- **Qwen-Image-2512-Fun-Controlnet-Union-2602**: Multi-condition ControlNet Union model (5 layer blocks) for the Qwen-Image-2512 architecture.
- **FLUX.1-dev-ControlNet-Union-Pro-2.0**: Next-generation unified 7-in-1 ControlNet for the FLUX.1-dev architecture by Shakker Labs.

---

## 📦 Available Models

| Filename | Base Architecture | Base Model | Supported Conditions | Quantization | File Size | License |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- |
| `Qwen-Image-2512-Fun-Controlnet-Union-2602_convrot_int8.safetensors` | Qwen-Image-2512 | [alibaba-pai/Qwen-Image-2512-Fun-Controlnet-Union](https://huggingface.co/alibaba-pai/Qwen-Image-2512-Fun-Controlnet-Union) | Canny, HED, Depth, Pose, MLSD, Scribble, Gray, Inpaint | ConvRot INT8 | ~1.64 GB | Apache-2.0 |
| `FLUX.1-dev-ControlNet-Union-Pro-2.0_convrot_int8.safetensors` | FLUX.1-dev | [Shakker-Labs/FLUX.1-dev-ControlNet-Union-Pro-2.0](https://huggingface.co/Shakker-Labs/FLUX.1-dev-ControlNet-Union-Pro-2.0) | Canny, Depth, Pose, Blur, Gray, Soft Edge, Low Quality | ConvRot INT8 | ~2.00 GB | Other / Non-Commercial |

---

## 🛠️ Model Features

- **Multi-Condition Control**: Unified architectures allowing single or combined conditioning inputs (Canny, Depth, Pose, etc.) without swapping model checkpoints.
- **ConvRot INT8 Precision**: Preserves high-frequency details and directional feature integrity through randomized orthogonal Hadamard rotations, eliminating outlier distortion common in standard INT8 quantization.
- **Significant Footprint Reduction**: Delivers ~50% to 70% VRAM and storage savings compared to unquantized FP16 checkpoints.

---

## 🚀 Usage in ComfyUI

To load and execute these ConvRot INT8 ControlNet models in ComfyUI, please use the dedicated loader node from the **ComfyUI-HSWQ-Loader-and-Tools** extension:

- **Extension Repository:** [ComfyUI-HSWQ-Loader-and-Tools](https://github.com/ussoewwin/ComfyUI-HSWQ-Loader-and-Tools)

### Installation

Clone the repository into your ComfyUI `custom_nodes` directory:

```bash
cd ComfyUI/custom_nodes
git clone https://github.com/ussoewwin/ComfyUI-HSWQ-Loader-and-Tools.git
```

Place the downloaded `.safetensors` files into your ComfyUI `models/controlnet/` directory and load them using the dedicated ControlNet loader node.

---

## 📜 Credits & License

### Base Models & Research
- **Qwen-Image ControlNet Union:** [alibaba-pai/Qwen-Image-2512-Fun-Controlnet-Union](https://huggingface.co/alibaba-pai/Qwen-Image-2512-Fun-Controlnet-Union) & [aigc-apps/VideoX-Fun](https://github.com/aigc-apps/VideoX-Fun) (Apache-2.0)
- **FLUX.1-dev ControlNet Union Pro 2.0:** [Shakker-Labs/FLUX.1-dev-ControlNet-Union-Pro-2.0](https://huggingface.co/Shakker-Labs/FLUX.1-dev-ControlNet-Union-Pro-2.0) & [InstantX Team](https://huggingface.co/InstantX) (FLUX.1-dev Non-Commercial License)
- **Base Architecture:** [black-forest-labs/FLUX.1-dev](https://huggingface.co/black-forest-labs/FLUX.1-dev)

---

**Disclaimer:** These models are provided for optimization, workflow acceleration, and research purposes. Please adhere to the licenses and terms of the respective base models.