Z-Image Logo
Z-Image.me
Back to List

Z-Image Turbo Fun ControlNet Union

Multi-functional ControlNet released by Alibaba PAI, supporting Canny, HED, Depth, Pose and MLSD controls

ControlNet
Image Control
Multi-functional
PAI
Visit Resource

Overview

This is a ControlNet model with 6 blocks added, trained from scratch on 1 million high-quality image dataset for 10,000 steps, supporting multiple control conditions.

Features

  • Supports Canny, HED, Depth, Pose and MLSD controls
  • 1328 resolution training
  • BFloat16 precision
  • control_context_scale adjustable control strength (0.65-0.80)
  • Detailed prompts improve stability

Images

Pose control input example
Generation result based on pose control
Canny edge detection input
Canny edge control generation result

Installation

# Clone VideoX-Fun repository
git clone https://github.com/aigc-apps/VideoX-Fun.git
cd VideoX-Fun
# Create model directories
mkdir -p models/Diffusion_Transformer
mkdir -p models/Personalized_Model
# Download weights to corresponding directories

Usage

Run examples/z_image_fun/predict_t2i_control.py

Requirements

  • Z-Image Turbo
  • VideoX-Fun environment
  • Additional preprocessing dependencies
  • GPU memory support