mabubaker1947/deeplearning repository overview

Setup Docker Container
When you have docker installed, you then need to 'pull' an instance of the image via commands below from the Docker website and you are all set to learn D E E P. To do this complete the following instructions:

Open a terminal window on your local machine.

In your terminal enter:

docker pull mabubaker1947/deeplearning:{tagname}

sudo docker run -it -p 8888:8888 -p 6006:6006 -v ~/:/host mabubaker1947/deeplearning:{tagname}

Enter your password if prompted (Note if you are a Windows user you might not require the sudo keyword in the above command).

Press Ctrl-d to exit from this image.

Note for Docker Toolbox. For point 1; there will be a Desktop link to "Docker Quickstart Terminal", click on this.

You should get a terminal with a message similar to the one below.

Docker is configured to use the default machine with IP 192.168.99.100

Accessing Your Version of The Docker Container The docker image should now be installed and a container setup on your local machine. To run the image in the container follow these instructions:

In your terminal enter: docker ps -a

A list (probably only with one entry) should appear with columns CONTAINER ID, IMAGE, etc. Copy the entry in the column CONTAINER ID which should be a string of letters and numbers.

In your terminal enter:

docker start YOUR_CONTAINER_ID

docker attach YOUR_CONTAINER_ID

You should now see the prompt of your terminal change. You should now be able to view the directories and files in the container. When you exit from the current Docker container, all changes made will be saved.

Launching Jupyter Notebook To open a Jupyter Notebook in the container, complete the following instructions:

In your terminal, enter:

sh run_jupyter.sh

Some output should appear. A web URL will also appear and should be something similar to http://0.0.0.0:8888/⁠. Copy this URL and enter it in a web browser. This should open the filesystem of the Docker container using Jupyter. Note for Docker Toolbox. In this case do not use the script above but type

jupyter notebook --allow-root

If a window does not automatically appear in the browser then open a new browser tab and type

http://0.0.0.0:8888/⁠

or

http://168.192.99.100:8888/⁠

NOTE:

Once you run a jupyter notebook, you can change the kernel in the file menu, Kernel->Change Kernel-> Python3.7.3 DL

(Python3.7.3 DL) is the kernel with most of the recent installations.

If you are using Docker Toolbox you will need to use the IP address from point 6 above, e.g "168.192.99.100". Also for Windows users you may need to use Google Chrome instead of Explorer/Edge.

Push an Update:

docker commit -m "commit message" -a "docker-image-name" dockerImageIDLocal "docker-Image-Name:TagName"

docker commit -m "Installed OpenCV" -a "mabubaker1947/deeplearning" 967a2a378h mabubaker1947/deeplearning:openCV

docker push mabubaker1947/deeplearning:openCV