This is a full-stack Django-based web application that integrates the OPENAI API by using the ChatGPT 3.5 Turbo and DALL-E 2 Models to generate random and unique artwork.

This project is live! Check it out here.

• Goals & MVP
• Tech Stack
• Build Steps
• Design Goals
• Project Features
• Additions & Improvements
• Learning Highlights
• Known Issues
• Challenges

The aim of this application is to create a full-stack application in the Django framework that can create random and unique images by leveraging modern tools including the OpenAI API.

• HTML
• CSS
• Django
• Python
• JavaScript
• DALL-E 2 API
• ChatGPT 3.5 Turbo API
• SQLite / Postgres
• AWS S3 Bucket
• Heroku

1. Clone the project from GitHub:

   git clone [email protected]:cyberforge1/imageForge.git
   

2. Register for an OpenAI API key at the following link.

3. Run the generateSecretKey.py file in the root directory to create a Django secret key.

4. Create a .env file in the root directory and attach the values of the keys generated in Steps 1 and 2:

   OPENAI_API_KEY= replace-this-value
   SECRET_KEY= replace-this-value
   

5. Create a virtual environment:

   python -m venv venv

6. Activate the virtual environment:

   • On Windows:

     venv\Scripts\activate

   • On macOS and Linux:

     source venv/bin/activate

7. Install project dependencies with pip:

   pip install -r requirements.txt

8. Set up and migrate the SQLite database:

   python manage.py migrate

9. Run the application locally:

   python manage.py runserver

To create an image, navigate to the generation page and click ‘generate’, then convert the unique prompt into unique art with a button click. The generated images can be viewed in a public gallery or a private history if logged in.

• This project was designed to be a fun and intuitive experience for users to navigate through the application to generate unique artwork.
• A careful selection of backgrounds throughout the application aims to create an immersive user experience.
• The Gallery and User History pages were inspired indirectly by the Instagram platform.

• A user registration and login system
• A collection of local data for creating randomized prompts
• A series of Python scripts that generate prompts and make API requests
• Generation of unique data using the ChatGPT 3.5 Turbo & DALL-E 2 Models
• A full-stack application to display this data

• Addition of delete functionality to the User History page
• Refine the Generation Page background image
• Improve styling on Login and Registration Pages
• Add a download link on modals for images
• Create mobile responsive design and mobile navbar

This project started as a passion project and was my first full-stack application

• Creation of a complete MVC backend
• Implementing API calls, handling the returned data and displaying it in the application
• Setting up local and remote databases
• Solidifying skills in Python scripting
• Deploying a Django application to Heroku

When a user is signed in and has a limited number of images generated, only a partial background image is displayed on the User History page.

• Creating a system that could store the images dynamically so that many users could access and interact with the application simultaneously. This was resolved by attaching an AWS S3 Bucket.
• Correcting the grammar and other inconsistencies of the locally generated prompts. This was resolved by making a call to the ChatGPT API.

• Visit my LinkedIn for more details.
• Check out my GitHub for more projects.
• Or send me an email at [email protected]