you can find the info here:PUBLIC REPOSITORY tensorflow/tensorflow
pull it:
docker pull tensorflow/tensorflow:latest
Start CPU only container
$ docker run -it -p 8888:8888 tensorflow/tensorflow
Go to your browser on http://localhost:8888/
- question1
When I enter this web page, it command me to input the password but I do not set it before. So I need to Setup a Password:
But where is the token value? you can find it in the shell log, just like:
- question2
in the jupyter I find that it only supports the python2 but I want the python3 kernel, so how to install it?
start a new terminal, with the command python3 --version I can find the python3 has been installed, but the pip3 is not by the command python3 -m pip --version.
Firstly install pip3,search the linux version with comand cat /proc/version,find it is linuxkit,then:
apt-get udpate
apt-get install python3-pip
python3 -m pip install ipykernel
python3 -m ipykernel install --user
but there is something wrong with my network and I can not find a way to make the python2 and python3 exist together, so I have to setup tensorflow again with the python3:
docker pull tensorflow/tensorflow:latest-py3
docker run -it -p 8888:8888 tensorflow/tensorflow:latest-py3
Start GPU (CUDA) container
Install nvidia-docker and run:
$ docker run --runtime=nvidia -it -p 8888:8888 tensorflow/tensorflow:latest-gpu
then Go to your browser on http://localhost:8888/
Other versions (like release candidates, nightlies and more)
See the list of tags. Devel docker images include all the necessary dependencies to build from source whereas the other binaries simply have TensorFlow installed.
For more details details see
https://github.com/tensorflow/tensorflow/blob/master/tensorflow/tools/docker/README.md
Reference
Support python3 on Docker image tensorflow
如何在 Docker 里切换 Python 版本