/generator9000

Web App for generating synthetic data

Primary LanguageTypeScriptBSD 3-Clause "New" or "Revised" LicenseBSD-3-Clause

🐈 Welcome to Generator 9000

✨ Your Starting Point for AI-Synthetic Data Generation

Start generating

Demo of Generator9000

Welcome to Generator9000! It's an open-source web app that's all about creating synthetic data objects tailored to your use case. We use Generative AI (GPT4 from OpenAI) to not only generate data objects with specific fields but also to create images based on these data objects. πŸš€

This tool is perfect if you're looking to build new, high-quality datasets for demos or Proof of Concept (PoC) projects, especially in these fast-evolving times. It's user-friendly, giving you full control over the prompts for both data and images, and other settings that directly affect the Large Language Models (LLM) you're working with.

You can tinker with all the settings, generate data objects, evaluate them, and then save your creations either on disk or in Weaviate. πŸ“βœ¨

Access Generator9000 here: https://www.gen9000.co/

πŸ”‘ You can enter your OpenAI key right in the frontend. It'll be saved as a httpOnly and secure cookie, making it super easy to pick up right where you left off next time you fire up the app. If you don't feel comfortable giving random websites your credentials, you can also host Generator9000 locally!

πŸš€ Getting Started

Before you dive in, make sure your machine is running Node v21.3.0 or higher. Here’s how you can get Generator9000 up and running:

  1. Clone the repository:

git clone https://github.com/weaviate/generator9000

  1. Navigate to the app directory:

cd generator9000_app

  1. Install the necessary packages:

npm install

  1. Create a .env.local file and add your OpenAI API key. If you want to use Weaviate, you can also add Weaviate variables:

OPENAI_API_KEY=YOUR_KEY WEAVIATE_URL=YOUR_URL WEAVIATE_API_KEY=YOUR_KEY

  1. Start the development server and start generating over 9000:

npm run dev

That's it! πŸŽ‰ Create, experiment, and have fun creating your very own synthetic datasets. Happy generating! 🌟