A simple tool utilizing the GPT-4-vision API to generate captions for images. This project is built with Gradio, and parts of the code were generated by GPT.
Developers: Jiaye and LEOSAM是只兔狲. Welcome everyone to add more new features to this project.
这是一款利用 GPT-4-vision API 为图像打标的简易工具。 使用 Gradio 构建，部分代码由 GPT 辅助生成。
Windows (If the automatic installation fails, please refer to the Manual Installation Instructions)
- Open Command Prompt as administrator and navigate to the directory where you want to clone the repository.
- Clone the repository using the following command:
git clone https://github.com/jiayev/GPT4V-Image-Captioner - Double-click
install_windows.batto run and install all necessary dependencies. - After the installation is complete, you can launch the GPT4V-Image-Captioner by double-clicking
install_windows.bat. - Hold down Ctrl and click on the URL in the terminal (or copy the URL to your browser), which will open the Gradio app interface in your default browser.
- Enter the official OpenAI or third-party GPT-4V API Key and API Url at the top of the interface. After setting the image address, you can start tagging the image.
- Open a terminal and navigate to the directory where you want to clone the repository.
- Clone the repository using the following command:
git clone https://github.com/jiayev/GPT4V-Image-Captioner - Navigate to the cloned directory:
cd GPT4V-Image-Captioner - Make the install and start scripts executable with the following command:
chmod +x install_linux_mac.sh; chmod +x start_linux_mac.sh - Execute the install script:
./install_linux_mac.sh - Launch the GPT4V-Image-Captioner in the terminal by executing the launch script:
./start_linux_mac.sh - Copy the URL displayed in the terminal and open it in your browser to access the Gradio app interface.
- Enter the official OpenAI or third-party GPT-4V API Key and API Url at the top of the interface. After setting the image address, you can start tagging the image.
Windows（如自动安装失败，请参考手动安装说明）
- 以管理员权限打开命令提示符，并导航到您想要克隆仓库的目录。
- 使用以下命令克隆仓库：
git clone https://github.com/jiayev/GPT4V-Image-Captioner - 双击
install_windows.bat运行，并安装所有必要的依赖项。 - 安装完成后，您可以通过双击
Start_windows.bat来在终端中启动GPT4V-Image-Captioner。 - 按住ctrl并点击终端中的URL地址（或复制URL地址在浏览器打开），将在默认浏览器中跳转打开Gradio应用界面。
- 请在界面最上方输入OpenAI官方或者第三方的GPT-4V API Key与API Url，设置图像地址后，就可以图像打标了。
- 打开终端，并导航到您想要克隆仓库的目录。
- 使用以下命令克隆仓库：
git clone https://github.com/jiayev/GPT4V-Image-Captioner - 导航到克隆的目录：
cd GPT4V-Image-Captioner - 使用以下命令使安装脚本和启动脚本变为可执行：
chmod +x install_linux_mac.sh; chmod +x Start_linux_mac.sh - 执行安装脚本：
./install_linux_mac.sh - 在终端中执行启动脚本来启动GPT4V-Image-Captioner。
./Start_linux_mac.sh - 复制终端中显示的URL地址，在浏览器中打开Gradio应用界面。
- 请在界面最上方输入OpenAI官方或者第三方的GPT-4V API Key与API Url，设置图像地址后，就可以图像打标了。
- One-Click Install & Launch: Added one-click installation (
install_windows.bat/install_linux_mac.sh) and launch (Start_windows.bat/Start_linux_mac.sh) features, simplifying the setup and startup process. - Environment Setup Instructions: Supplemented the installation and launch instructions for the program under Windows and Linux environments.
- Speed Optimization: Improved the speed of tagging images. Now annotations can be completed within 2-3 seconds.
- Tag Management: Provided different handling options for images with existing tags: "Overwrite", "Prepend", "Append", and "Skip".
- Subfolder Processing: The new program can now process all image files within folders and subfolders, supporting formats like '.png', '.jpg', '.jpeg', '.webp', '.bmp', '.gif', '.tiff', '.tif'.
- Interruption Feature: Added the ability to interrupt the batch tagging process.
- Error Screening: Images that fail GPT tagging (e.g., NSFW content) can be moved to a new folder based on keywords.
- Localization: Added support for the Chinese language.
- 一键安装和一键启动: 增加了一键安装 (
install_windows.bat/install_linux_mac.sh) 和一键启动 (Start_windows.bat/Start_linux_mac.sh) 功能，简化了程序的安装和启动过程。 - 环境说明补充: 补充了在Windows和Linux环境下程序的安装和启动说明。
- 运行加速: 提高了程序的打标速度。现在可以在2-3秒内完成一张图片的标注。
- 标签处理: 对于已有标签的图像文件，提供了以下不同处理选项："覆盖", "前置插入", "结尾追加" 和 "跳过"。
- 子文件夹处理: 新程序能够处理文件夹及其子文件夹中的所有图像文件，支持的图像格式包括：'.png', '.jpg', '.jpeg', '.webp', '.bmp', '.gif', '.tiff', '.tif'。
- 程序中断: 增加了在批量打标签过程中中断打标的功能。
- 报错筛查: 可以根据关键词，将所有GPT标记失败的图像（例如NSFW内容）移动到新的文件夹中。
- 本地化: 增加了对中文的支持。
-
Open the Command Prompt by pressing
Win + R, typingcmd, and then pressingEnter. -
Clone the repository to your local machine using the following command:
git clone https://github.com/jiayev/GPT4V-Image-Captioner -
Once cloning is complete, navigate to the cloned directory:
cd GPT4V-Image-Captioner -
Before installing any dependencies, make sure that Python is installed on your system. Check for Python's presence by typing the following command and pressing
Enterin the Command Prompt:python --versionIf Python is not installed, you will get an error message. In that case, please visit the Python official download page and follow the instructions to install it.
-
Create a virtual environment named
myenvto avoid contaminating the global Python environment:python -m venv myenv -
Activate the virtual environment you just created:
myenv\Scripts\activate -
Update
pipto date:python -m pip install --upgrade pip -
Install the
requests,gradio, andtqdmlibraries within the virtual environment:pip install scipy networkx wordcloud matplotlib Pillow tqdm gradio requests -
After completing the steps above, you can start GPT4V-Image-Captioner by double-clicking the
Start_windows.batfile.
-
按
Win + R打开命令提示符。键入cmd然后按Enter。 -
使用下面的命令克隆仓库至本地：
git clone https://github.com/jiayev/GPT4V-Image-Captioner -
克隆完成后，切换到克隆的目录中：
cd GPT4V-Image-Captioner -
在安装依赖库之前，在命令提示符中输入以下命令并按
Enter来检查是否电脑已经安装了 Python：python --version如果未安装，会显示错误信息。请访问 Python 官方下载页面 并按照指示进行安装。
-
创建一个名为
myenv的虚拟环境以避免污染全局 Python 环境：python -m venv myenv -
激活你刚创建的虚拟环境：
myenv\Scripts\activate -
更新
pip至最新版本：python -m pip install --upgrade pip -
在虚拟环境中安装
requests、gradio、tqdm等库：pip install scipy networkx wordcloud matplotlib Pillow tqdm gradio requests -
完成上述步骤后，可通过双击
Start_windows.bat文件来启动 GPT4V-Image-Captioner。