add: gradio webui for train and use

README.md

@@ -190,6 +190,57 @@ Installation using Docker, docker-compose as follows:
 1. Directly download the [officially built installation package](https://github.com/GuijiAI/HeyGem.ai/releases)
 2. Double-click `HeyGem-x.x.x-setup.exe` to install
 
+### WebUI Usage Guide
+
+HeyGem provides a simple and intuitive interface to help you train and use digital human models (Please make sure you have completed the server-side deployment on Windows):
+
+#### Environment Requirements
+
+- Python dependencies: Ensure you have installed the necessary Python libraries
+  ```
+  gradio==4.44.1
+  requests~=2.32.3
+  pydub==0.25.1
+  uuid==1.30
+  python-json-logger==3.3.0
+  ```
+
+#### Language Switching Feature
+
+HeyGem supports Chinese and English interface switching:
+- Default launch with Chinese interface: `python app.py`
+- Launch with English interface: `python app.py --lang en`
+- Launch with Chinese interface: `python app.py --lang zh`
+
+#### Data Storage Paths
+
+- Audio files storage path: `D:\heygem_data\voice\data`
+- Video files storage path: `D:\heygem_data\face2face\temp`
+
+#### Training Digital Humans
+
+1. After opening the software, switch to the "Digital Human Training" tab
+2. Upload a reference video: Click on the "Upload Reference Video" area to select a video file with a person speaking
+3. Enter a name for the digital human: Name your digital human model
+4. Click the "Start Training" button
+5. Wait for the training to complete; upon success, the message "Training Successful!" will be displayed
+
+#### Digital Human Synthesis
+
+1. Switch to the "Digital Human Synthesis" tab
+2. Select a trained digital human model from the dropdown menu
+3. Choose a synthesis method:
+   - **Text Input**: Enter text content and click the "Synthesize" button
+   - **Audio Upload**: Upload an audio file and click the "Synthesize" button
+4. The system will assign a task ID and begin processing
+5. Click the "Query Synthesis Status" button to check the progress
+6. Once synthesis is complete, the video result will be displayed in the interface
+
+#### Other Features
+
+- **Refresh Digital Human Model List**: Click this button to update the dropdown menu of trained models
+- To view storage paths and other detailed information, expand the "Storage Path Information" area at the top of the interface
+
 ## Open APIs
 
 We have opened APIs for model training and video synthesis. After Docker starts, several ports will be exposed locally, accessible through `http://127.0.0.1`.

README_zh.md

@@ -190,6 +190,57 @@ Heygem是一款专为Windows系统设计的全离线视频合成工具，它能
 1. 直接下载[官方构建的安装包](https://github.com/GuijiAI/HeyGem.ai/releases)
 2. 双击`HeyGem-x.x.x-setup.exe`即可安装
 
+### WebUI使用说明
+
+HeyGem提供了简单直观的界面，帮助您训练和使用数字人模型（请确保您已经在Windows完成了服务端部署）：
+
+#### 环境要求
+
+- Python依赖：确保安装了必要的Python库
+  ```
+  gradio==4.44.1
+  requests~=2.32.3
+  pydub==0.25.1
+  uuid==1.30
+  python-json-logger==3.3.0
+  ```
+
+#### 语言切换功能
+
+HeyGem支持中英文界面切换：
+- 默认启动为中文界面: `python app.py`
+- 启动英文界面: `python app.py --lang en`
+- 启动中文界面: `python app.py --lang zh`
+
+#### 数据存储路径
+
+- 音频文件存储路径：`D:\heygem_data\voice\data`
+- 视频文件存储路径：`D:\heygem_data\face2face\temp`
+
+#### 数字人训练
+
+1. 打开软件后，切换到"数字人训练"标签页
+2. 上传参考视频：点击"上传参考视频"区域选择一个包含人物说话的视频文件
+3. 输入数字人名称：为您的数字人模型命名
+4. 点击"开始训练"按钮
+5. 等待训练完成，成功后将显示"训练成功!"消息
+
+#### 数字人合成
+
+1. 切换到"数字人合成"标签页
+2. 从下拉菜单中选择已训练的数字人模型
+3. 选择合成方式：
+   - **文字输入**：输入文字内容，点击"合成"按钮
+   - **音频上传**：上传音频文件，点击"合成"按钮
+4. 系统会分配一个任务ID并开始处理
+5. 点击"查询合成状态"按钮查看进度
+6. 合成完成后，视频结果将显示在界面上
+
+#### 其他功能
+
+- **刷新数字人模型列表**：点击此按钮可更新已训练模型的下拉菜单
+- 如需查看存储路径等详细信息，可展开界面上方的"存储路径信息"区域
+
 ## 开放 API
 
 我们开放了模特训练和视频合成的API，Docker 启动后会在本地暴露几个端口，通过`http://127.0.0.1`可以调用。