deep-learning-getting-started

Setting up Tensorflow and deep learning resources

Setting up Tensoflow

Installation

Follow the Docker installation instructions
Make sure to test your installation
In your working directory create a folder named notebooks to save all your notebooks developped using Tensorflow.

Launching Tensorflow

Launch your image using the following command (assuming you are using the latest-devel image):

docker run -it -p 8888:8888 -p 6006:6006 -v $PWD/notebooks:/root/notebooks gcr.io/tensorflow/tensorflow:latest-devel

-p 8888:8888: exposes the port 8888 (Jupyter Notebooks) -p 6006:6006: exposes the port 6006 (Tensorboard) -v $PWD/notebooks:/root/notebooks: mounts a volume to save and reuse your notebooks between sessions.

Launching Tensorboard

tensorboard --logdir=/root/notebooks/[LOG DIRECTORY]

Resources