Public Repository

Last pushed: 5 months ago
Short Description
Same usage as dash00/tensorflow-python3-jupyter, add tflearn package
Full Description

Exactly same usage as dash00/tensorflow-python3-jupyter,just add tflearn package.

What does it contain ?

  • git
  • Jupyter Notebook
  • TensorFlow
  • tflearn
  • scikit-learn
  • pandas
  • matplotlib
  • numpy
  • scipy
  • Pillow
  • Python 2 and 3

How to install it ?

Download the environment

$ docker pull shuang/tensorflow-tflearn-python3-jupyter

How to use it ?

Once Docker has been installed and the package has been downloaded, one can simply use the following commands from a terminal (use Docker Quickstart Terminal on Windows).

Let <CONTAINER_IP> be the IP address of the container

  • on Windows: note the IP address which is printed in the terminal at the top (here: 192.168.99.100 on the picture above)
  • on Linux: you will use 'localhost' or 127.0.0.1
  1. Basic usage: Hello world!

    Use the following command to start a basic container

     $ docker run -it -p 8888:8888 dash00/tensorflow-python3-jupyter
    

    The options -it and -p allow respectively to run an interactive container (attached to the terminal) and to expose the port 8888 of the container (this port is used by the jupyter web service).

    By default, a token authentification is enabled. Therefore, the first time you use the container you will need to copy the token from the terminal to log in the Jupyter Notebook (see next subsection to disable authentification).

    Then, you can access your notebooks from your web browser at this URL (paste the token in the password field) :
    http://<CONTAINER_IP>:8888/

    Open the hello world notebook, run it and voila !

  2. Disable token authentification

    Even if it is not recommended for security reasons, the token authentification can be disabled. To remove it, you will have to explicitely call the run_jupyter.sh script with the option --NotebookApp.token set to empty:

     $ docker run -it -p 8888:8888 dash00/tensorflow-python3-jupyter /run_jupyter.sh --allow-root --NotebookApp.token=''
    
  3. Use a persistent folder

    If you want to work in persistent folder (independent of the container, which will not be removed at the end of the container execution) use the -v option as follow:

     $ docker run -it -p 8888:8888 -v /$(pwd)/notebooks:/notebooks dash00/tensorflow-python3-jupyter
    

    You can change /$(pwd)/notebooks by any path on the local system. If the folder does not exist, it will be created. This option maps the given local folder with the folder of the notebooks on Jupyter. This folder should contain all your notebooks indeed.

  4. Use Jupyter Notebook and Tensorboard in the same time

    a. Create a container 'notebooks' to run Jupyter Notebook (port 8888)

     $ docker run  --name notebooks -d -v /$(pwd)/notebooks:/notebooks -v /$(pwd)/logs:/logs -p 8888:8888 dash00/tensorflow-python3-jupyter /run_jupyter.sh --allow-root --NotebookApp.token=''
    

    The option -d detaches the container, i.e. it makes it run in background. Jupyter is still available on http://<CONTAINER_IP>:8888/.

    b. Create a container 'board' to run Tensorboard (port 6006):

     $ docker run  --name board -d -v /$(pwd)/logs:/logs -p 6006:6006 dash00/tensorflow-python3-jupyter tensorboard --logdir /logs
    

    Tensorboard will be available on http://<CONTAINER_IP>:6006/.

How to build my own docker image ?

使用 Docker 快速配置深度学习(Tensorflow)环境

Docker Pull Command
Owner
shuang0420