The following editor will bring you an article on TensorFlow installation and jupyter notebook configuration methods. The editor thinks it’s pretty good, so I’ll share it with you now and give it as a reference. Let’s follow the editor to take a look.

tensorflow uses anaconda to install under ubuntu, jupyter notebook running directory and remote access configuration

Installing Anaconda under Ubuntu

bash ~/file_path/file_name.sh

After permission appears, you can press Ctrl+C to skip, yes agree.

After the installation is completed, you will be asked whether to add the path path, or you can modify the file content yourself

Close the command console and reopen it

python -V

You can view it Whether the installation is successful

Modify the python version of anaconda to meet the tf requirements

conda install python=3.5

Anaconda installs TensorFlow

Get Anaconda's TF source

anaconda search -t conda tensorflow

After waiting, the list will be displayed. I selected the conda-forge/tensorflow source to install

anaconda show conda-forge/tensorflow

will display the current installation information and prompt the instructions that need to be entered to install: conda install --channel XXXX URL

After input, it will prompt for the package information that needs to be installed, enter Y Agree and install.

The download speed of some packages is very slow, you can open lantern first

jupyter notebook usage and configuration

Enter jupyter under Ubuntu Open the notebook

jupyter notebook --generate-config to generate the default configuration file jupyter_notebook_config.py

Default address configuration

Modify the configuration file c.NotebookApp.notebook_dir = 'XXXX'

Remote access

No remote access by default

First generate the login password: open ipython and enter

from notebook.auth import passwd

passwd( ) Enter the password to be set according to the prompts, and copy the generated sha1 password

which needs to be copied completely, for example:

sha1:ce23d945972f:34769685a7ccd3d08c84a18c63968a41f1140274

Modify the configuration file

c.NotebookApp.ip='*'
c.NotebookApp.password = u'sha:xxxxxxxxxxxxxxxxxxxxxxxxxx'
c.NotebookApp.open_browser = False
c.NotebookApp.port =8888

Note that there is * in the ip, and there is u

in front of the password. After the configuration is completed, it can be accessed remotely. After the configuration is completed, it will not pop up automatically when starting the notebook. The webpage needs to be opened manually. After opening, you need to enter the password before you can access it.

Note 1: Ubuntu can use ifconfig to obtain the local IP address

Note 2: The virtual machine can use NAT mode or bridge mode to make the host accessible to the notebook

The above is the detailed content of TensorFlow installation and detailed explanation of jupyter notebook configuration. For more information, please follow other related articles on the PHP Chinese website!