Using Docker via nvidia-docker2

The main installation page is here:

https://github.com/NVIDIA/nvidia-docker  

Some concept guides are discussed here:

https://chunml.github.io/ChunML.github.io/project/Installing-NVIDIA-Docker-On-Ubuntu-16.04/

https://hub.docker.com/r/tensorflow/tensorflow/ 

The arguments for using Docker and Tensorflow are given here:

https://www.quora.com/Is-there-any-advantage-of-running-Tensorflow-in-a-docker-container-rather-than-just-pip-installing-it

Additional References:

https://github.com/tensorflow/tensorflow/tree/master/tensorflow/tools/docker 

https://www.tensorflow.org/serving/docker 

https://github.com/nvidia/nvidia-docker/wiki/Installation-(version-2.0)

https://becominghuman.ai/how-to-build-a-deep-learning-server-based-on-docker-bea70b8bd2c7  

https://stackoverflow.com/questions/25185405/using-gpu-from-a-docker-container 

https://medium.com/@ceshine/docker-nvidia-gpu-nvidia-docker-808b23e1657 

https://hackernoon.com/docker-compose-gpu-tensorflow-%EF%B8%8F-a0e2011d36

https://towardsdatascience.com/how-to-deploy-machine-learning-models-with-tensorflow-part-2-containerize-it-db0ad7ca35a7