GraphExecuter

</div> <div align="center">

中文 | English

</div>

GraphExecuter is an open-source software developed based on NodeGraphQt, adopting a multi-threaded graph execution architecture. Users only need to write node logic and construct graphs by connecting nodes, and the system will automatically schedule and execute complex workflows.

This system combines visual node editing with efficient parallel computing, making it suitable for scenarios such as data processing, automated tasks, and distributed computing. Developers do not need to worry about thread management—they can focus solely on implementing node functionalities to quickly build high-performance workflows.

GraphExecuter provides a convenient solution for users who require flexible, scalable, and visual orchestration, catering to diverse needs from experimentation to production.

code: https://github.com/laoxue888/GraphExecuter

update:

• 2025-7-28：docker容器开发方式
• 2025-7-5: 【开源】测试运行yolov13实时检测算法
• 2025-4-17: 开源DeepSeek人工智能语音对话工作流，你的AI智能语音助手啦
• 2025-4-2: 高效创建工作流，可实现类似unreal engine的蓝图效果，内部使用多线程高效执行节点函数

Quick Start

❇️ Install

conda create -n graph_executer python=3.12
conda activate graph_executer

cd graph_executer
pip install -r .\requirements.txt

pip install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 --index-url https://download.pytorch.org/whl/cu124

cd graph_executer/downloads
git clone https://github.com/laoxue888/NodeGraphQt.git
cd NodeGraphQt
pip install -e .

❇️Run

python main.py

Related Projects

• NodeGraphQt
• NodeGraphQt Documentation

改进方向

• 添加导航栏和多窗口。
• 节点的运行模式可以选择，比如可从节点图的首个节点开始运行，或者从最后一个节点运行，或者循环运行，或者只运行一次。
• 做一个节点代码编辑的窗口ui。
• 有些后端服务使用FastAPI。
• 那个节点在运行，该节点的颜色就发生变化。