Conda-Jupyter-Docker

• How to build own docker container from Anaconda docker image?
• How to create Conda environment from a .yml file in the docker container?
• How to add the conda environment to the kernel list in Jupyter notebook?
• How to launch jupyter notebook from the container?

Access the Jupyter notebook from your remote machine over SSH

$ ssh -L 2222:localhost:2222 <REMOTE_USER>@<REMOTE_HOST>

Note: Also you can access to remote machine without jupyter notebook as following:

$ ssh <REMOTE_USER>@<REMOTE_HOST>

Go to your docker path

$ /Path/to/working/directory/docker 

Create a Dockerfile

# Install anaconda
FROM continuumio/anaconda

# Set working directory for the project
WORKDIR /Path/to/working/directory/

# Install packages to new conda environment
COPY environment.yml .
RUN conda env create -f environment.yml

# Set PATH environment variables
ENV PATH /opt/conda/envs/env_name:$PATH  

Build docker container

Note: No need to rebuild once built!

$ docker build --no-cache -t deep_learning_docker .

Run docker container

$ HOST_PORT=${2222}
$ CONTAINER_PORT=${8888}
$ docker run -it --rm --gpus all -p $HOST_PORT:$CONTAINER_PORT -v /Path/to/working/directory/:/Path/to/working/directory/ deep_learning_docker

Note that: -it defines interactive terminal, -v bind mount a volume, --gpus all allow all GPU devices to add to the container and -p 2222:8888 publish a container's port(s) to the host.

Activate conda environment

$ source activate env_name

After executing the above command we will be in the environment

(env_name) root@############:/Path/to/working/directory/#

Add your environment to the kernel list in jupyter notebook

If reveive No module named ipykernel

$ pip install ipykernel 

Note: No need to rebuild once built!

$ python -m ipykernel install --user --name=env_name

Launch Jupyter notebook with default port from docker container

$ jupyter notebook --ip=0.0.0.0 --port=8888 --no-browser --allow-root

Enter the URL in your local browser

To find hostname:

$ hostname -I
>> HOSTNAME=${127.0.0.1}

Open a browser from your local machine and navigate to http://${HOSTNAME}:2222/tree or 127.0.0.1:2222/?token=${TOKEN} and enter your token specified in your terminal.