PyTorch Docker image

Ubuntu + PyTorch + CUDA (optional)

This respository is a fork of anibali/pytorch/

Requirements

In order to use this image you must have Docker Engine installed. Instructions for setting up Docker Engine are available on the Docker website.

If you have a CUDA-compatible NVIDIA graphics card, you can use a CUDA-enabled version of the PyTorch image to enable hardware acceleration. I have only tested this in Ubuntu Linux.

Firstly, ensure that you install the appropriate NVIDIA drivers and libraries. If you are running Ubuntu, you can install proprietary NVIDIA drivers from the PPA and CUDA from the NVIDIA website.

You will also need to install nvidia-docker2 to enable GPU device access within Docker containers. This can be found at NVIDIA/nvidia-docker.

Features

We have three types of docker files:

Type	Feature
dockerfile-runtime	The lightweight dockerfile which does not support CUDA compile
dockerfile-devel	The heavyweight dockerfile which supports cuda compile, use it if you need to compile packages with CUDA
dockerfile-apex	You will need that if you only want to build the packages once and need a lightweight docker image

Build images

1. Modify the environment.yml to add your own packages, and move it to the directory of docker file.

2. Build the docker images using the following command:

$ docker pull muyeby/docker-pytorch:torch1.4

Prebuilt images

Pre-built images are available on Docker Hub under the name muyeby/pytorch. For example, you can pull the CUDA 10.0 version Pytorch1.4 with:

$ docker pull muyeby/docker-pytorch:torch1.4

The table below lists software versions for each of the currently supported

CUDA	PyTorch
no-cuda	1.2.0
cuda-11.3	1.10.0, 1.12.0
cuda-11.1	1.8.1
cuda-10.2	1.5.1, 1.6.0, 1.8.1
cuda-10.1	1.3.1, 1.4.0
cuda-10.0	1.0.1, 1.1.0, 1.2.0
cuda-9.2	0.4.1
cuda-9.1	0.4.0
cuda-9.0	0.3.0, 0.3.1, 0.4.1
cuda-8.0	0.2.0, 0.3.0, 0.3.1
cuda-7.5	0.3.0