NOTE: Please rotate the CUDA repo keys as mentioned here: https://forums.developer.nvidia.com/t/notice-cuda-linux-repository-key-rotation/212772 and pull the container from NGC: https://catalog.ngc.nvidia.com/orgs/nvidia/containers/pytorch

NVDIA's GPU-optimized distribution of PyTorch.
PyTorch is a GPU accelerated tensor computational framework with a Python front end. Functionality can be easily extended with common Python libraries such as NumPy, SciPy and Cython.

This image bundles NVIDIA's container for PyTorch into the NGC base image for AWS. The NGC base image is an optimized environment for running the GPU optimized containers for Deep Learning and HPC available from the NGC (NGC) container registry. The Docker containers available on the NGC container registry are tuned, tested, and certified by NVIDIA to take full advantage of NVIDIA GPU's.

This image allows user to immediately start using PyTorch . You can learn more about the PyTorch distribution from NVIDIA and look for newer releases of NVIDIA's PyTorch container at http://ngc.nvidia.com

This 23.03.0 image release contains:
Ubuntu Server 20.04
NVIDIA Driver 470.103.01
Docker-ce 20.10.12
NVIDIA Container Toolkit 1.8.1
NVIDIA Container Runtime 3.8.1
NVIDIA's GPU-optimized PyTorch container 22.02-py3