Walk through art gallery of your own choice and flavor with paintings on arbitrary topic (within the guidelines of OpenAI GPT-3 and DALLE 2 models 🌝).
You can test the demo here by creating a temporary API token. It will create 12 512x512 image generations with DALLE model and around 300-600 tokens of text-davinci-003 model per one run. The token is only stored on client side, you can check the sources for proof. Also, you can checkout the video here.
To run locally first set your OpenAI API key in config.ts file, see config.example.ts for reference. Then run the following commands
yarn
yarn serve
and enjoy your own gallery.
- Download created pieces
- Really infinite corrdior that updates as you move
- A few other cool ideas