NVIDIA NGC Tutorial: Run a PyTorch Docker Container on Ubuntu with Lambda Stack.
Full blog post: https://lambdalabs.com/blog/set-up-a-tensorflow-gpu-docker-container-using-lambda-stack-dockerfile/
This tutorial shows you how to install Docker with GPU support on Ubuntu Linux. To get GPU passthrough to work, you’ll need docker, nvidia-container-toolkit, Lambda Stack, and a docker image with a GPU accelerated library.
How it works:
1) Lambda Stack installs the drivers, CUDA, docker, and nvidia-container-toolkit
2) NVIDIA NGC provides a container with GPU enabled PyTorch
Note, that you can run containers created with your own dockerfiles using this method as well — I’ve just shown it with NGC because it’s already hosted. If you want to make your own docker images with Lambda Stack inside, you can use our open source dockerfiles here:
https://github.com/lambdal/lambda-stack-dockerfiles
ubuntu
