# BertSafeChat

**Repository Path**: echo-wind-dev/bert-safe-chat

## Basic Information

- **Project Name**: BertSafeChat
- **Description**: 基于Bert-Base-Chinese作为预训练模型的恶意评论识别系统
- **Primary Language**: Unknown
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-04-07
- **Last Updated**: 2025-04-07

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# BertSafeChat - 中文攻击性言论检测系统

[English Version Below]

## 📖 项目简介

基于BERT的中文攻击性言论检测系统，提供Flask API和命令行两种使用方式。模型使用`bert-base-chinese`微调，可有效识别常见网络暴力、歧视性言论。

## 🚀 主要特性

- 支持实时文本检测（API模式）
- 提供命令行交互界面
- 基于Transformer的高精度模型
- 低延迟推理（支持GPU加速）
- 完善的错误处理机制

## 📦 项目结构

```
.
├── flask_api/               # Flask API相关代码
│   └── app.py               # API主程序
├── model/                   
│   └── best_model.bin       # 预训练模型文件
├── main_cmd.py              # 命令行交互程序
├── LICENSE                  # Apache 2.0许可证
├── README.md                # 项目说明文件
└── .gitignore               # Git忽略规则
```

## ⚙️ 安装指南

### 环境要求
- Python 3.8+
- PyTorch 1.8+
- CUDA 11.0+（如需GPU加速）

### 安装步骤
```bash
# 克隆仓库
git clone https://gitee.com/your-username/bert-safe-chat.git

# 安装依赖
pip install -r requirements.txt

# 下载预训练模型（可选，已包含best_model.bin可跳过）
# 将best_model.bin放置在model目录下
```

## 🛠 使用方式

### API模式
```bash
# 启动API服务
cd flask_api
python app.py

# 测试请求
curl -X POST http://localhost:5000/predict \
  -H "Content-Type: application/json" \
  -d '{"text": "测试文本"}'
```

### 命令行模式
```bash
python main_cmd.py
>> 请输入待检测文本：今天天气真好
>> 检测结果：正常言论
```

## 🤝 贡献指南

欢迎通过Issue或Pull Request参与项目贡献！  
1. Fork本仓库
2. 创建特性分支（git checkout -b feature/AmazingFeature）
3. 提交修改（git commit -m 'Add some AmazingFeature'）
4. 推送分支（git push origin feature/AmazingFeature）
5. 发起Pull Request

## 📄 开源协议

本项目采用 [Apache License 2.0](LICENSE)

---

# BertSafeChat - Chinese Offensive Speech Detection System

## 📖 Project Introduction

BERT-based Chinese offensive speech detection system, providing both Flask API and command-line interface. The model is fine-tuned with `bert-base-chinese`, effectively identifying common cyber violence and discriminatory speech.

## 🚀 Key Features

- Real-time text detection (API mode)
- Command-line interface
- High-accuracy Transformer-based model
- Low-latency inference (GPU acceleration supported)
- Robust error handling

## 📦 Project Structure

```
.
├── flask_api/               # Flask API source
│   └── app.py               # API main program
├── model/                   
│   └── best_model.bin       # Pretrained model
├── main_cmd.py              # CLI program
├── LICENSE                  # Apache 2.0 License
├── README.md                # Documentation
└── .gitignore               # Git ignore rules
```

## ⚙️ Installation

### Requirements
- Python 3.8+
- PyTorch 1.8+
- CUDA 11.0+ (for GPU acceleration)

### Installation Steps
```bash
# Clone repo
git clone https://gitee.com/your-username/bert-safe-chat.git

# Install dependencies
pip install -r requirements.txt

# Download pretrained model (optional)
# Place best_model.bin in model directory
```

## 🛠 Usage

### API Mode
```bash
# Start API service
cd flask_api
python app.py

# Test request
curl -X POST http://localhost:5000/predict \
  -H "Content-Type: application/json" \
  -d '{"text": "Sample text"}'
```

### Command-line Mode
```bash
python main_cmd.py
>> Input text: Test input
>> Result: Normal speech
```

## 🤝 Contributing

Contributions are welcome via Issues or Pull Requests!  
1. Fork the Project
2. Create your Feature Branch (git checkout -b feature/AmazingFeature)
3. Commit your Changes (git commit -m 'Add some AmazingFeature')
4. Push to the Branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

## 📄 License

Distributed under the [Apache License 2.0](LICENSE)