TARS Terminal Chatbot

TARS is a terminal-based chatbot inspired by the character TARS from Interstellar. It leverages modern LLMs, tool integrations, and persistent memory to deliver engaging and context-aware interactions.

Features

• Interactive Chat: Chat with TARS in your terminal, complete with wit and useful responses.
• Tool Integration: Includes search capabilities using APIs (e.g., Tavily).
• Persistent Memory: Maintains context between interactions using a memory module.
• Configurable Personality: Modify the system prompt to fine-tune TARS's personality and behavior.

Prerequisites

• Python 3.8 or higher
• pip for installing dependencies
• API keys for Tavily and Groq (or other LLM APIs)

Installation

1. Clone the repository:

   git clone https://github.com/Rikhil-Nell/Terminal-Chatbot.git
   cd tars-terminal-chatbot

2. Set up a virtual environment (optional but recommended):

    python3 -m venv venv
    source venv/bin/activate  # On Windows, use venv\Scripts\activate

3. Install the required dependencies:

   pip install -r requirements.txt

4. Create a .env file in the project root with your API keys:

   GROQ_API_KEY=your_groq_api_key
   TAVILY_API_KEY=your_tavily_api_key

5. Modify the system prompt if desired:

   • Update the prompt.txt file to define TARS's personality and tone.

Usage

Run the chatbot using:

    python app.py

Once running, you'll see a terminal interface where you can chat with TARS. Enter your messages, and TARS will respond with contextually aware and detailed answers.

File Structure

core.py: Main logic of the chatbot. app.py: Sets up the terminal interface for inference. prompt.txt: Defines the system prompt for TARS's personality. requirements.txt: List of dependencies. .env: Configuration file for storing API keys (not included in the repository).

Configuration

• System Prompt: Customize prompt.txt to adjust TARS's behavior.
• Memory Management: Persistent memory ensures TARS remembers previous interactions. Adjust the memory settings in tars_chatbot.py as needed.

Contributions

Contributions are welcome! If you'd like to improve TARS or add features, please submit a pull request.

License

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