Docker container with JupyterLab

This dockerfile contains instruction to build container with Jupyter from Anaconda.
Also, there is GPU support for TensorFlow, PyTorch, XGBoost, LGBM.
Additional packages: OpenCV, CatBoost

Build

Use this command for build:

sudo docker build -t name:tag <path/to/Dockerfile>

(this may take a lot of time)
Or just download it from docker hub (~12 GB):

sudo docker pull voudy/jupyter

Run

If you want to run with GPU support use:

sudo nvidia-docker run --rm -it -v <local/dir/for/notebooks>:/working_dir -p <local port>:8888 name:tag

But you can run it only on CPU:

sudo docker run --rm -it -v <local/dir/for/notebooks>:/working_dir -p <local port>:8888 name:tag

After that go to localhost:port or other ip you would like to use.

For downloaded version use voudy/jupyter instead of name:tag