A GAN (Generative Adversarial Network) using PyTorch

SETTING UP NEW ENV FOR TORCH

Steps for setting up conda env to avoid conflicts in packaging while setting up PyTorch:

conda create --name gan

conda activate gan

conda install -c pytorch pytorch=2.0 (The latest version of PyTorch) Or

conda install pytorch torchvision torchaudio -c pytorch

We may get this error: RemoveError: 'packaging' is a dependency of conda and cannot be removed from conda's operating environment

If that is the case, then use conda update --force conda

on your new conda env and try again.

Once all packages are installed, we install matplotlib in this new environment: conda install matplotlib jupyter

Before opening Jupyter Notebook, we need to register the conda gan environment so that we can create Notebooks using it as the kernel. To do that, with the gan environment activated, we run: python -m ipykernel install --user --name gan

Now we can open Jupyter Notebook by running on the shell:

jupyter notebook