Notebook

Abstract :

A container image of jupyter notebook development environment with anaconda3, some other runtimes and packages.

Runtimes :

• python3 (Anaconda3-2020.11)
• python2
• octave
• c

Packages :

• tensorflow==2.4.0
• pytorch==1.7.0
• pyspark==2.4.5
• pyod==0.7.5

Quick start :

cd /path/to/your/directory/
docker run -d --rm -p 8888:8888 --name notebook -v $PWD:/opt/notebooks mooyxu/notebook
# Then Open your browser with url `<ip>:8888` for notebook
#                          or url `<ip>:8888/lab` for lab.

Usage :

1. GET IMAGE (Optional, but suggested)

    docker pull mooyxu/notebook

2. ALIAS (Optional, but suggested)

   • LINUX

     echo alias notebook=\''docker run -d --rm -p 8888:8888 --name notebook -v $PWD:/opt/notebooks mooyxu/notebook'\' >> ~/.bashrc
     source ~/.bashrc

   • or MAC

     echo alias notebook=\''docker run -d --rm -p 8888:8888 --name notebook -v $PWD:/opt/notebooks mooyxu/notebook'\' >> ~/.bash_profile
     source ~/.bash_profile

3. RUN

   • WITHOUT PASSWORD

     cd /path/to/your/directory/
     notebook
     # Depend on step 2, otherwise use command 'docker run -d --rm -p 8888:8888 --name notebook -v $PWD:/opt/notebooks mooyxu/notebook'.

   • or WITH PASSWORD

     docker run -it --rm mooyxu/notebook python -c 'from notebook.auth import passwd; print(passwd())'  # Input PASSWORD, copy OUTPUT, then paste it below
     cd /path/to/your/directory/
     notebook /bin/bash /opt/run.sh '<OUTPUT>'  # [EXAMPLE] notebook /bin/bash /opt/run.sh 'a1b2c3'
     # Depend on step 2, otherwise use command: docker run -d --rm -p 8888:8888 --name notebook -v $PWD:/opt/notebooks mooyxu/notebook /bin/bash /opt/run.sh '<OUTPUT>'.

4. USE

   Open your browser with url <ip>:8888 for notebook

   or url <ip>:8888/lab for lab.

5. STOP

    docker stop notebook

6. REMOVE

    docker rmi mooyxu/notebook \
    | docker images | grep mooyxu/notebook | awk '{print $3}' \
    | xargs docker rmi \
    && sed -i '/mooyxu\/notebook/d' ~/.bashrc  # or MAC: instead of '~/.bash_profile'

ADVANCED :

1. BUILD

    docker build -t mooyxu/notebook .
    # Suggested in linux: docker build -t mooyxu/notebook -t mooyxu/notebook:`date -I` .

2. DEBUG

    docker run -it --rm -p 8888:8888 --name notebook -v $PWD:/opt/notebooks mooyxu/notebook bash
    jupyter notebook
    # Then you will get one or two images: mooyxu/notebook:latest and mooyxu/notebook:<date_of_today>

3. TEST tensorflow

    import tensorflow as tf;print(tf.reduce_sum(tf.random.normal([5000, 5000])))

4. TEST pytorch

    from __future__ import print_function
    import torch
    x = torch.rand(5, 3)
    print(x)

5. QUICK commands

    alias pip='docker exec -it notebook pip'
    alias ipython='docker exec -it notebook ipython'
    alias pyspark='docker exec -it notebook pyspark'
    alias java='docker exec -it notebook java'
    # alias octave='docker exec -it notebook octave'

Contact :

Email: xumengyu2008@qq.com