GPT CLI

This command-line interface (CLI) tool enables the use of LLM models directly from your terminal. It is designed to be cost-effective by utilizing the latest "turbo" versions of the GPT models, including GPT 4 turbo, and Anthropic Claude as well.

Edit Dec 2023: I made a terrible mistake; the original way to invoke this cli tool was gpt <prompt>, but it never occured to me that gpt is an existing tool--GUID partition table maintenance utulity. Silly me.

Table of Contents

• Usage
• Prerequisites
• Configuration
• Installation

Usage

% prompt "Hello, GPT3.5-turbo!"
Hello! How can I assist you today?
==================================
Model: gpt-3.5-turbo-1106 | Token usage: 27 | Cost: 0.0001 USD

% prompt -v 4 "Hello, GPT4-turbo!"
Hello! How can I assist you today?
==================================
Model: gpt-4-1106-preview | Token usage: 27 | Cost: 0.0011 USD

Images speak louder than words.

Prerequisites

• Python 3.9 or higher (needed to create a virtual environment)
• Python Pip

Installation

cd install_scripts
chmod +x install.sh
source ./install.sh .bashrc (or .zshrc)

Changing the invocation alias

To change the alias, go to invoke_gpt.sh and locate the function:

prompt() {
    source $GPT_APP_PATH/venv/bin/activate
    python3 $GPT_APP_PATH/src/app.py "$@"
    deactivate
}

Change this function name (prompt) to whatever you'd like.

Lastly, re-source your .zshrc or .bashrc to activate the invocation alias