Simple command line tool for text to image generation using OpenAI's CLIP and Siren (Implicit neural representation network). Technique was originally created by https://twitter.com/advadnoun
First off Im pretty new to this project. I've been trying to generate different results with deep-daze, but with the default settings it is set to take on average 24 hours per epoch. When I was generating I happened to check 'radeontop' to see how much VRAM deep-daze was using, and it seemed to only be using 1gb, which is what background ubuntu normally runs at anyway. When I check system monitor it shows that most if not all my cores for my CPU are running at near 100%. So it seems that deep-daze is somehow using my CPU instead of GPU??? Would anyone know how to fix this?
System:
Ryzen 5 2600x six core x 12
Radeon RX 590 (8gb VRAM)
24gb RAM
First off Im pretty new to this project. I've been trying to generate different results with deep-daze, but with the default settings it is set to take on average 24 hours per epoch. When I was generating I happened to check 'radeontop' to see how much VRAM deep-daze was using, and it seemed to only be using 1gb, which is what background ubuntu normally runs at anyway. When I check system monitor it shows that most if not all my cores for my CPU are running at near 100%. So it seems that deep-daze is somehow using my CPU instead of GPU??? Would anyone know how to fix this?
System: Ryzen 5 2600x six core x 12 Radeon RX 590 (8gb VRAM) 24gb RAM
Ubuntu 20.04.2 LTS