Public | Automated Build

Last pushed: 4 months ago
Short Description
Personal Docker Image for Deep Learning in Python.
Full Description

Python Deep Learning


This repository is based off


I followed these tutorials to get started with Cuda, cudnn and nvidia-docker.

Docker Image

The docker image can be found here: . It can also be built from the docker file running make build BUILD=build



The Makefile assumes that the data will be in {$HOME}/deeplearning/data and will map this to /data inside the Docker container. It also assumes that the Jupyter notebooks (or the source code) is located in {$HOME}/deeplearning/notebooks. This can all be changed when calling make.


These commands will pull the latest image from dockerhub by default ( If you want it to build the docker file and use the built image instead, append BUILD=build to any of these commands (also, for now I recommend setting TAG=local to avoid name collision when you pull).

Jupyter notebook

make notebook [DATA=<path-to-data-directory>] [SRC=<path-to-src-directory>]


make ipython [DATA=<path-to-data-directory>] [SRC=<path-to-src-directory>]


make bash [DATA=<path-to-data-directory>] [SRC=<path-to-src-directory>]

Executing a bash session in a running container

# Get the container name
$ docker container ls
$ docker container exec -it <container-name> bash
Docker Pull Command