Artist-Engineering-Geek 0.1.0

A bunch of GANs and data downloaders to make a custom AI artist

Artist

Motivation

• An easy to edit codebase for Progressive GAN originally published by this research paper and other GANs
• Supplement my personal projects
• Resume builder
• For fun and to understand state-of-the-art AI

Installation

CUDA Installation

Ensure you have a GPU if you want to train in any reasonable amount of time.

• Install CUDA here
• Install cuDNN here

Project Installation

pip install Artist-Engineering-Geek
# Don't forget to install your specific pytorch and torchvision libraries for your gpu
# in my case, I have the NVIDIA RTX 3090 so this is my version
pip install --no-cache-dir --pre torch torchvision torchaudio -f https://download.pytorch.org/whl/nightly/cu112/torch_nightly.html

Running the program

To train the program, run the "Train.ipynb" notebook and alter your parameters at will on GitHub. There should be sufficient in-code documentation for you to understand what the hell is going on