MLDS docker environment for machine learning with GPU support.
Follow the steps and turn an empty instance to a ready-to-go deep learning environment in 30 mins.
-
0-prepare-host.sh
install docker, nvidia driver, nvidia-docker in your host machine. -
1-build-mlds.sh
build machine learning toolbelt container
here I choose Tensorflow gpu supported version for deep learning, you may append whatever you like in Dockerfile! -
2-start-mlds.sh
start the container with some environment setup script, with ipython notebook in http://<your_machine>:8888