BUILD.md

Building the containers

The main build script can be executed with:

bash build.sh

It builds Dockerfiles (found in the src folder) into container images in the following order:

Dockerfile	Container	Remarks
base.Dockerfile	nvaitc/ai-lab:YY.MM-base	Ubuntu+Conda+Jupyter+PyData
tf2.Dockerfile	nvaitc/ai-lab:YY.MM-tf2	++ TensorFlow 2.x
full.Dockerfile	nvaitc/ai-lab:YY.MM	++ PyTorch, RAPIDS
vnc.Dockerfile	nvaitc/ai-lab:YY.MM-vnc	++ VNC Desktop, VirtualGL

Adding additional packages into a container

You can use any of the containers as a base to build your own containers. The entire historical directory of containers can be accessed on Docker Hub.

Building your own container

As a simple example, to add the MXNet framework into container:

# create and build this Dockerfile

FROM nvaitc/ai-lab:20.03
LABEL maintainer="You <you@yourdomain.com>"

# you need to use root user for apt-get or make install
#USER root
#RUN apt-get update && apt-get install some-package

# use notebook user for pip/conda
USER $NB_UID
RUN pip install --no-cache-dir mxnet-cu101mkl

# always switch back to notebook user at the end
USER $NB_UID

After which, we can build the Dockerfile:

docker build . -f mxnet.Dockerfile -t myCustomMxNet:latest

Where -f points to the Dockerfile and -t defines the name of the output container.