fritzprix / tf_lab

Docker builder to play with Tensorflow & scikit-learn

for those who are learning ML with old books

Steps

1. build docker image

   $ ./build.sh

2. run jupyter notebook

   $ ./run.sh

local directory binding

there are 3 sub-directories mounted to docker for convenience. each of them is intended for..

• work : repository for .ipynb files
• tf_logs : data exported from tensorflow (e.g. models, logs, etc.)
• data : sample data for training of your own

modules included

• python2
• Tensorflow (1.15 non-GPU)
• pandas
• numpy
• matplotlib
• scikit-learn