- Ubuntu 14.04
- Python 2.7
- Ngrok
- Tmux
- CUDA 8
- cuDNN v5
- Tensorflow
- Theano
- Keras
- iPython/Jupyter Notebook (including iTorch kernel)
- Numpy, SciPy, Pandas, Scikit Learn, Matplotlib
You need a working install of Nvidia-Docker
docker run \
-it \
--rm \
-p 222:22 \
-p 8888:8888 \
jmizell/ml-python:latest
- SHELL_PASSWORD: specify the ssh password for the root user
- NGROK_AUTH: ngrok auth token. If this variable is set ngrok will expose the ssh daemon on a public tunnel. Check the container log for this value.
docker build -t ml-python .