How to use Docker, GPU on Linux

Ubuntu 18.04

1. Install Docker: https://docs.docker.com/install/

$ sudo apt install docker.io

2. Install NVIDIA docker support: https://github.com/NVIDIA/nvidia-docker

# Add the package repositories
$ distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
$ curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
$ curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.list
$ sudo apt-get update && sudo apt-get install -y nvidia-container-toolkit
$ sudo systemctl restart docker

3. Grant permission

$ sudo setfacl -m user:pengy6:rw /var/run/docker.sock

4. Download a TensorFlow Docker image

$ docker pull tensorflow/tensorflow:1.15.0-gpu

$ docker pull pytorch/pytorch

5. Start a TensorFlow Docker container

$ docker run -it --rm pytorch/pytorch bash

6. Mount local folder (in order)

$ docker run --mount type=bind,source=source,target=target  -it --rm pytorch/pytorch bash