TensorFlow using Docker
So you are ready to burn some code and Model on Docker.
This time you want to try TensorFlow on Docker. You can write code for TensorFlow on Java, C++, Go and Python. But most preferred language is python and C++. You can use editors like Atom and sublime. Sublime is faster than atom but both are having good features to support TensorFlow.
I can assume already you have installed docker in your Windows or Linux or Mac Machine. If not, i have other articles which will help you to install docker on your machine.
So lets add TensorFlow image to your Local Docker
docker run -it gcr.io/tensorflow/tensorflow bash
This installation is required drive space is at about 1GB.
After the above command gets executed successfully, then following should be the screen you should expect. Just for your reference, i am using MAC, but you can use any OS like Windows or Linux.
Now next is, we need to test, whether we have install it correctly or not. To do we need to run a small python script. The image that we are running on docker is already have required python pre-configured so we dont need to install anything.
At the end you should expect message
Hello, TensorFlow !