Skip to content

Latest commit

 

History

History
55 lines (37 loc) · 1.92 KB

README.md

File metadata and controls

55 lines (37 loc) · 1.92 KB
Raw

ChatGPT-cli

A simple python script for interfacing with ChatGPT in the command prompt.

Usage

chatgpt-cli.py --help

chatgpt-cli.py
Usage:  chatgpt-cli.py --[option]
Options:
        --help                          Display this help message.
        --key <keyfile>                 Specify an api key file. (Default behavior is to get this from $OPENAI_API_KEY.)
        --prompt "prompt"               Print reply from prompt and exit. (Default behavior is to run in interactive mode.)
        --max-history <number>          Set maximum chat memory for chatbot. (Doing this will decrease the size of api calls.)
        --log <logfile>                 Set a log file to save the chat history in.
        --model "model"                 Specify which model to use. (Defaults to gpt-4o-mini.)
        --system-message "message"      Provide the chatbot with some context or instructions about its behavior.

Dependencies

The following are dependencies that you may need to install.

  • requests
  • prompt_toolkit

Installation / Configuration

Installation

  1. Clone this git repository.

    git clone https://github.com/joshuadeal/chatgpt-cli

  2. Change to repository directory.

    cd ./chatgpt-cli/

  3. Install dependencies.

    python -m pip install -r ./requires.txt

Configuration

  1. You can create a json file with the value "apiKey" and pass it to the program each time you run it via the "--key" option, but this method is not recommended.

    Said json file should look something like this...

     {
         "apiKey": "Put your API key here."
 }

  2. Your best bet is to set up the environmental variable "OPENAI_API_KEY"

    1. echo "export OPENAI_API_KEY='put your api key here'" >> ~/.bashrc"

    2. source ~/.bashrc

    To test if this has been done correctly run echo $OPENAI_API_KEY. The output should be your API key.

Take care to keep API keys secure!