Visual Art
Digital Art
~1 min read
Deep Haze Generative Art
These images were made using Deep Haze, the simple command line tool for text to image generation using OpenAI's CLIP and Siren. Each image was generated from a text prompt with the following hyperparameters: a learning rate of 1e-5, 16 hidden layers, 8 epochs, 256 loss calculations per epoch, 4 samples per iteration, and 4 generated images per loss calculation.
Adobe Fresco
These images were made using Adobe Fresco, the powerful drawing app with the world's largest digital brush collection.

















