GPT-4 CLI Chatbot

This project contains a command-line interface (CLI) chatbot powered by OpenAI's GPT-4 model. It allows users to interact with the GPT-4 model in a conversational manner directly from the terminal.

Features

• Real-time interaction with the GPT-4 model.
• Colored prompts for better user experience.
• Ability to stream the model's response back to the user.
• Configurable settings via an external JSON file.

Installation

Ensure you have the necessary libraries installed using pip:

pip install -r requirements.txt

Configuration

1. Create a .env file in the project root and add your OpenAI API key like so:

OPENAI_API_KEY=your-api-key-here

2. Edit the config.json file in the project root with the following structure:

{
    "messages": [
        {
            "role": "system",
            "content": "You are a helpful assistant."
        }
    ],
    "response_params": {
        "temperature": 0.3,
        "stream": true
    }
}

• The messages array contains the initial messages to set the behavior of the model.
• The response_params object contains parameters for the openai.ChatCompletion.create method to control the behavior of the model.

Usage

Initial Setup

Create virtualenv and install requirements.

Create a .env and populate it with your OPENAI_API_KEY=

Running the Script

Using Python directly:

python main.py

or:

Using a Global Shortcut (bash): To make the script globally accessible, create a symlink to it in /usr/local/bin. This allows you to use the chat command from any directory. Make sure to run this command as sudo to get the necessary permissions:

sudo ln -s /absolute/path/to/workspace/gpt-cli/scripts/run.sh /usr/local/bin/chat

Important: Remember to chmod +x run.sh to make the script executable.

Run as sudo, sudo chat.

Using Powershell:

Edit $projectPath in run.ps1 to match your environment.

Add this line to your $PROFILE:

Set-Alias -Name chat -Value "C:\your\directory\here\workspace\gpt-cli\scripts\run.ps1"

Then use 'chat' to run from anywhere.

You'll be greeted with a prompt (>>>). Here, you can type your message to the GPT-4 model. The model's response will be streamed back to you in real-time.

Contributing

Feel free to fork this repository, make changes, and submit pull requests. Any contributions are welcome!