jmackles / inspectigator

Proof of concept of leveraging multiple openai calls to validate expected results, or to reject them.

Inspectigator

Inspectigator is a proof of concept Python application that leverages multiple OpenAI API calls to validate expected results or reject them. It allows users to input a color and shape, which then generates a corresponding image using DALL·E 3. The generated image is further inspected by GPT-4's vision capabilities to ensure the image contains the specified color and shape.

Features

• User input validation for color and shape.
• Image description generation using OpenAI's GPT-4 model.
• Image creation with DALL·E 3 based on the generated description.
• Image inspection using GPT-4's vision capabilities to validate the content.
• Local storage of generated images.
• Direct opening of generated image URLs in a web browser.

Prerequisites

Before you begin, ensure you have met the following requirements:

• You have installed Python 3.9 or higher.
• You have an OpenAI API key with access to GPT-4 and DALL·E 3 models.
• You have installed the required Python packages listed in requirements.txt.

Installation

To install Inspectigator, follow these steps:

1. Clone the repository:

   git clone https://github.com/Jmackles/inspectigator.git

2. Navigate to the project directory:

   cd inspectigator

3. Install the required packages:

   pip install -r requirements.txt

Usage

To use Inspectigator, run the following command in your terminal:

python inspectigator.py

Follow the prompts to enter a color followed by a shape (e.g., "red circle"). The program will attempt to generate and validate an image up to three times before exiting.

Configuration

Create a .env file in the root directory of the project with the following content:

OPENAI_API_KEY=your_openai_api_key_here

Replace your_openai_api_key_here with your actual OpenAI API key.