A Gentle Introduction to Deep Learning with TensorFlow

Michelle Fullwood

Slides in Reveal.js for the talk above, to be presented at PyCon 2017 in Portland, OR.

Jupyter

Below are the instructions for running the JupyterLab notebook.

Development

To deploy a simple single user environment (i.e. for development) simply run the following Docker commands (note: this assumes you are already in the Deep-Learning-With-TensorFlow-PyCon2017 directory):

docker run -d \
           --rm \
           --name gidltf \
           -e JUPYTER_ENABLE_LAB=yes \
           -p 8888:8888 \
           -v $PWD/notebooks:/home/jovyan \
           jupyter/tensorflow-notebook:lab-3.4.5 && \
sleep 5 && \
docker logs gidltf 2>&1 | grep "http://127.0.0.1" | tail -n 1 | awk '{print $2}'

Click the link (should look similar to: http://127.0.0.1:8888/lab?token=LONG_ALPHANUMERIC_STRING) which will automatically log you in and allow you to start running the notebooks.