Telegram Chatbot with Message Summarization

This project is a Telegram chatbot that can interact with users, maintain chat history, and provide summaries of past conversations. The bot is designed to be flexible, allowing it to operate in either a local mode or via an API, depending on the configuration.

demo.mp4

Features

• Chat History Management: The bot keeps track of recent messages in a chat, allowing it to provide context-aware responses.
• Message Summarization: Users can request summaries of the last n messages, with options for both English and Chinese summaries.
• Superchat Support: The bot can handle messages in different threads within a Superchat, maintaining separate histories for each thread.
• Customizable System Prompt: The bot's behavior can be customized by modifying the system prompt in the configuration file.
• Local and API Modes: The bot can operate in two modes:
  • API Mode (recommended): Connects to an external API that supports OpenAI API (e.g., DeepSeek) for generating responses.
  • Local Mode: Uses a local chatbot model (Qwen2.5-7B-Instruct) for generating responses.

Installation

API Mode (Recommended)

1. Clone the Repository:

   git clone [email protected]:runqiuw/context_chatbot_telegram.git
   cd telegram-chatbot

2. Install Dependencies: Ensure you have Python 3.8 or higher installed. Then, install the required packages:

   pip install json
   pip install logging
   pip install python-telegram-bot --upgrade
   pip install openai

3. Configure the Bot:

   • Create a config.json file in the root directory with the following structure:

     • To begin, you need an Access Token. If you haven't got one, follow the steps here.

     {
       "TOKEN": "YOUR_TELEGRAM_BOT_TOKEN",
       "username": "YOUR_BOT_USERNAME",
       "mode": "api",  
       "api_key": "YOUR_API_KEY", 
       "base_url": "YOUR_API_BASE_URL", 
       "model": "YOUR_MODEL_NAME",
       "context_length": 100
     }

   • Replace the placeholders with your actual bot token, username, and API details. See the Configuration section below for more details.

4. Run the Bot:

   python main.py

Local Mode

1. Clone the Repository:

   git clone [email protected]:runqiuw/context_chatbot_telegram.git
   cd telegram-chatbot

2. Install Dependencies:

   • Ensure:
     1. You have Python 3.8 or higher installed.
     2. You have an Nvidia GPU with at least 8GB display memory.
     3. You have installed pytorch that is compatible with your CUDA. If not, use nvcc -V to check the version of your CUDA and install the corresponding pytorch version from the official website.
     4. Your hard drive has at least 15GB of free space.
   • Then, install the required packages:
     • If you have trouble insntalling autoawq, this may help.

   pip install json
   pip install logging
   pip install python-telegram-bot --upgrade
   pip install transformers -U
   pip install autoawq

3. Configure the Bot:

   • Create a config.json file in the root directory with the following structure:

     • To begin, you need an Access Token. If you haven't got one, follow the steps here.

     {
       "TOKEN": "YOUR_TELEGRAM_BOT_TOKEN",
       "username": "YOUR_BOT_USERNAME",
       "mode": "local", 
       "api_key": "YOUR_API_KEY",
       "base_url": "YOUR_API_BASE_URL",  
       "model": "YOUR_MODEL_NAME",  
       "context_length": 100 
     }

   • Replace the placeholders with your actual bot token, username, and API details. See the Configuration section below for more details.

4. Run the Bot:

   python main.py

Configuration

The bot's behavior can be customized by modifying the config.json file:

• TOKEN: Your Telegram bot token.
• username: The username of your bot.
• mode: Set to "local" to use a local chatbot model or "api" to use an external API.
• api_key: The API key for the external service (required if mode is "api").
• base_url: The base URL for the external API (required if mode is "api").
• model: The model name to use with the external API (required if mode is "api").
• context_length: The maximum number of messages to store in the chat history.

Usage

• Start a Conversation: Simply mention the bot in a message (e.g., @yourbotname Hello!), and it will respond.
• Clear Chat History: Use the /new_chat command to clear the chat history and start fresh.
• Summarize Messages: Use the /summarize <number> <(optional)zh/en> command to get a summary of the last n messages. You can specify the language (zh for Chinese, en for English) as an optional argument.
• Help: Use the /help command to get a list of available commands and their usage.

Contributing

Contributions are welcome! Please feel free to submit issues or pull requests.

1. Fork the Repository.
2. Create a New Branch (git checkout -b feature/YourFeatureName).
3. Commit Your Changes (git commit -m 'Add some feature').
4. Push to the Branch (git push origin feature/YourFeatureName).
5. Open a Pull Request.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Acknowledgments

• Thanks to the python-telegram-bot library for providing the tools to interact with the Telegram API.
• Special thanks to DeepSeek for providing the API used in this project.
• Thanks to Qwen for the open-source model Qwen2.5-7B-Instruct

Enjoy chatting with your new Telegram bot! 🚀