Using pytorch#

../../_images/pytorch-logo-dark.png

PyTorch is the framework and Python interface on top of the torch library, which has been specifically built for machine learning purposes.

| Website | GitHub | Documentation |

The origins of torch go back to the IDAP and EPFL research institutes. PyTorch has originally been built at Meta AI.

PyTorch mainly offers two things. A tensor computing library following the numpy api closely, and Deep Neural Networks

The tensor computing library works both with a CPU and a GPU backend and can therefore easily be used to help us overcoming the single-core computing boundary.