This repository gathers the code and information about NFT art generation with GANs - Generative adversarial networks.
The standard in NFT collections is usually of 10.000 single items. I've been using the artwork of Camelia Otero, an artist from my hometown in the south of France, to train my GANs. If you wish to keep going with the challenge of generating similar artwork than the one of Camelia Otero, you can access 1000 images of her paintings at this google drive folder. Images have been collected by scraping the instagram page oterocamelia with the authorisation from the artist.
I am basically asking my GAN models to reproduce the style of Camelia's paintings, which it doesn't achieve yet but I can still get some interesting results with some of them.
A presentation of the early results is available here