For CPU Bound ML Research
For a GPU computing container - See this dockerhub page and this github repo.
This image is built to run on CPU-only compute, which you may want to do if you need high CPU density. For example, I use usually use X1 instance types on AWS.
See all running containers:
docker ps -a -f status=running
Run terminal attached to a running container:
docker exec -it <container_name> bash
Save state of container
docker commit <container_id> new_image_name:tag_name(optional)
More Resources
See this tutorial