Блокнот Jupyter из контейнера Docker localhost не работает - PullRequest
Новая
       6

Блокнот Jupyter из контейнера Docker localhost не работает

0 голосов
/ 28 марта 2020

Я пытаюсь запустить ноутбук Jupyter из Docker контейнера, и localhost просто не запускается в моем браузере.

Вот мой скрипт для запуска jupyter notebook 

#!/bin/bash

# Usage:
# ./run_jupyter.sh [image]
#
# Parameters:
#   image - optional. A custom image to use instead of default. If it is given, 
#           we will not attemp to pull the latest image. This allows local image development.
#
# Container will be removed upon exit, but any jupyter settings and pakcages installed with
# pip install --user are kept in mlcourse.ai/home folder. 
# Remove them if you need a completely fresh image.
#
IMAGE=${1:-festline/mlcourse_ai}

#custom port
PORT=4545

# If we are called without a parameter to specify a new image, 
# let's make sure we are on the latest image
if [ $# -eq 0 ]; then
    echo "Will check for the latest image on the docker hub."
    docker pull $IMAGE
fi

exec docker run --rm -u $(id -u):$(id -g) -v "$PWD:/notebooks" -w /notebooks -e HOME=/notebooks/home -p $PORT:8888 $IMAGE jupyter-notebook --NotebookApp.ip=0.0.0.0 --NotebookApp.password_required=False --NotebookApp.token='' --NotebookApp.custom_display_url="http://localhost:$PORT"

# this allows for container to be created and persisted.
# which means that you can keep the changes you made, 
# i.e. if you installed more software with pip.
# 
# USER_ID=$(id -u) GROUP_ID=$(id -g) GROUP_NAME=$(id -gn) exec docker-compose -f docker/docker-compose.yaml up

Когда я запускаю скрипт, это то, что он показывает на моем терминале 

[W 07:14:08.593 NotebookApp] All authentication is disabled.  Anyone who can connect to this server will be able to run code.
[I 07:14:08.629 NotebookApp] JupyterLab extension loaded from /opt/conda/lib/python3.6/site-packages/jupyterlab
[I 07:14:08.629 NotebookApp] JupyterLab application directory is /opt/conda/share/jupyter/lab
[C 07:14:08.631 NotebookApp] Running as root is not recommended. Use --allow-root to bypass.
...