safari-tensorflow-05MAY2017

Dockerfile file for Tensorflow Course (15 May 2017)

Use this Dockerfile to build a container with course material and required libraries installed.

Following are the steps to build and use the container.

1. Clone this repo

2. CD into the cloned folder

3. Run docker build -t <your tag> .

4. Once successfully built you can run container with this command:

   docker run -p 127.0.0.1:88:8888 -d --name tensorflow <your tag>

We now need the security/login key from jupyter notebook to start using it. To get the key run this command:

docker logs tensorflow

This will print the log output from machine like:

[C 01:57:48.215 Tensorflow]
Copy/paste this URL into your browser when you connect for the first time,
to login with a token: http://localhost:8888/?token=2bc2971f5327aa018bf591621e5b97cd98b741ce2c6b5743

Now navigate to http://127.0.0.1:88 You will be asked for the token. Copy the string after token= from the log output above and paste into the textbox, click on the Login button and you are good to go.

Note:- If you do not specify a tag (<your tag>) in step 3, you will need to note/copy the id of the image created and pass it in place of the tag later.

##Alternate

If you don't want to build the image, you can pull the one I pushed to the docker hub. Use this command to get image from the hub:

docker pull amantur/safari-tensorflow-15may2017:v1

Once image is downloaded you can follow the directions above from step 4 onwards.