Docker container for CSCI 4850/5850 - Neural Networks (ver. 2022-12-10)
This container is built on top of jupyter/datascience-notebook provided by jupyter/docker-stacks. It provides a JupyterLab environment with several essential (and non-essential) tools used in CSCI4850/5850 - Neural Networks.
The recommended way to obtain the docker image is to pull from DockerHub:
docker pull jlphillips/csci4850:2022-12-10
docker run -it --rm -p 8888:8888 --user root -e JUPYTER_ENABLE_LAB=yes -e GRANT_SUDO=yes -v /home/jphillips:/home/jovyan/work jlphillips/csci4850:2022-12-10
You will need to modify /home/jphillips to instead indicate where your files are located in order to make this work...
You will also need to make sure your Docker installation is set up to use the Nvidia container toolkit, as described here, or just sudo apt-get install nvidia-container-toolkit if you are using NVIDIA/CUDA on your host.
If you want to build the image yourself (rather than pulling the image from DockerHub - note this is probably not what you want to do), you may do the following...
To prep:
git clone https://github.com/jlphillipsphd/CSCI4850.git
To build:
docker build -t csci4850 CSCI4850
To run:
docker run -it --rm -p 8888:8888 --user root -e JUPYTER_ENABLE_LAB=yes -e GRANT_SUDO=yes -v /home/jphillips:/home/jovyan/work csci4850