Jupyter notebook

From Deep Learning Course Wiki
Jump to: navigation, search

We use Jupyter Notebook for nearly all of our analysis. If you haven't used it before, you may like to try a tutorial, such as [1], [2], or [3]. Jupyter notebook used to be called 'ipython notebook', so you will still often see this terminology. Don't let that confuse you - the two terms refer to the same thing!

If you installed your GPU server using one of our Installation scripts, then your Jupyter server is already running on port 8888 on your server, so just go to server-name:8888 in your browser to access it. The password we have set up for you is dl_course (that is 'dl' as in '[D]eep [L]earning').

Common Problems

  • No execution of a cell: The symptom is a blinking "Connected" sign on the top right corner and no cell is executed and an asterisk stays still in the square bracket. The remedy is to run jupyter notebook --port=8889 instead of jupyter notebook in your command line. [4]