用于批量图片的标签推断工具，基于 WD14 ONNX 模型与 tags.csv，输出与图片同名的 .txt 标签文件。提供图形界面，支持阈值调整、排除标签、递归目录处理与失败处理策略。

• 批量处理图片（jpg/png/webp/bmp/gif）并生成标签文件
• 文件与文件夹拖拽输入
• 实时显示处理速度与剩余时间（ETA）
• 通用/角色阈值可调，支持评分与角色标签输出
• 支持下划线替换为空格、排除标签列表（支持按添加顺序/名称排序，重复检测，多套配置）
• 支持递归子目录、跳过失败或已存在输出
• 自动保存设置到 settings.json

• Windows
• Python 3.10+（源码运行）
• 依赖库：PyQt6、onnxruntime、Pillow、numpy
• 模型与标签文件：WD14 onnx 模型与 tags.csv

1. 运行程序：

   python batch_tagger.py

2. 在界面中填写：

   • 输入文件夹
   • 输出文件夹（为空则使用输入目录）
   • 模型（onnx）路径
   • 标签（tags.csv）路径
   • 可选的 ComfyUI 目录

3. 设置参数并点击“开始”。

• 每张图片输出同名的 .txt 文件，写入推断得到的标签
• 输出目录不存在时会自动创建

• settings.json：保存界面输入与参数设置
• startup_error.log：启动异常记录
• diagnostics.json：诊断信息
• debug.log：调试日志

使用 PyInstaller 的 spec 文件进行打包：

pyinstaller BatchTagger.spec

• batch_tagger.py：主程序
• BatchTagger.spec：PyInstaller 打包配置
• comfyui-WD14-Tagger/：ComfyUI WD14 Tagger 依赖资源

BatchTagger is a powerful GUI application for batch tagging images using ONNX Runtime and WD14 Tagger models. It allows users to efficiently tag large collections of images with high accuracy and customizable settings.

• Batch Processing: Tag multiple images in a directory recursively.
• High Performance: Optimized startup and processing speed using ONNX Runtime.
• Smart Filtering: Real-time validation for exclude tags to prevent duplicates.
• User-Friendly Interface: Easy-to-use GUI for setting thresholds, excluded tags, and model selection.
• Responsive Layout: Modern grid layout for settings to maximize screen real estate.
• Cross-Platform: Built with Python and Qt (PyQt6), compatible with Windows, Linux, and macOS.

• Python 3.10 or higher
• Git

1. Clone the repository:

   git clone https://github.com/madlaxcb/batchtagger.git
   cd batchtagger

2. Create a virtual environment (optional but recommended):

   python -m venv venv
   # Windows
   venv\Scripts\activate
   # Linux/macOS
   source venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

   Note: Ensure you have a compatible ONNX Runtime installed for your hardware (CPU/GPU).

python batch_tagger.py

Download the latest release from the Releases page. Extract the archive and run BatchTagger.exe.

1. Select Directory: Choose the folder containing your images.
2. Configure Settings: Adjust thresholds for general and character tags.
3. Exclude Tags: Add any tags you want to exclude from the output.
4. Start Tagging: Click the "Start" button to begin processing.

To build a standalone executable using PyInstaller:

1. Install development dependencies:

   pip install pyinstaller

2. Run the build command:

   pyinstaller BatchTagger.spec

The executable will be located in the dist/BatchTagger/ directory.

Contributions are welcome! Please feel free to submit a Pull Request.

MIT License