Skip to content

Latest commit

 

History

History

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 

ReadMe.md

Docker setup instructions for TensorPy

1. Install the Docker Toolbox from https://www.docker.com/products/docker-toolbox

2. Create your for TensorPy Docker environment:

docker-machine create --driver virtualbox tensorpy
(If your Docker environment ever goes down for any reason, you can bring it back up with a restart.)
docker-machine restart tensorpy

3. Configure your shell:

eval "$(docker-machine env tensorpy)"

4. Go to the TensorPy home directory on the command line, which is where Dockerfile is located. (This assumes you've already cloned the TensorPy repo.)

5. Create your Docker image from your Dockerfile: (Get ready to wait awhile)

docker build -t tensorpy .

6. Run the example test

docker run tensorpy ./run_docker_test.sh

7. You can also enter Docker and stay inside the shell:

docker run -i -t tensorpy

8. Now you can run an example test from inside the Docker shell:

./test_classify_image.sh

9. When you're satisfied, you may exit the Docker shell:

exit

10. (Optional) Since Docker images and containers take up a lot of space, you may want to clean up your machine from time to time when they’re not being used:

http://stackoverflow.com/questions/17236796/how-to-remove-old-docker-containers Here are a few of those cleanup commands:

docker images | grep "<none>" | awk '{print $3}' | xargs docker rmi
docker rm 'docker ps --no-trunc -aq'

If you want to completely remove all of your Docker containers and images, use these commands: (If there's nothing to delete, those commands will return an error.)

docker rm -f $(docker ps -a -q)
docker rmi -f $(docker images -q)

Finally, if you want to wipe out your TensorPy Docker virtualbox, use these commands:

docker-machine kill tensorpy
docker-machine rm tensorpy

11. (Optional) More reading on Docker can be found here: