I have set up two servers with more powerful CPUs and GPUs running the Tensorflow software we will need for this course. I have also set up a user account for everyone in the class on each machine. To log in to either of these machines you will use your Lawrence network user name in all caps as your user name and your Lawrence ID number as your password.
These are the network addresses for the two machines:
143.44.10.62 cmsc106.net
You will be able to access these machines and run Jupyter notebooks from inside Visual Studio Code on your laptop.
The first thing you will need to do is to install a new extension. Click the extentions tab in VS Code and search for the SSH extension. Install this extension.
Once you have installed the extension you should see a little colored square that looks like >< in the lower left hand corner of the VS Code window. To connect to one of our servers, start by clicking on that square. This will cause a list of commands to show up in the command palette. Select the Connect to Host command, and then type in your login information. For example, I type
GREGGJ@143.44.10.62
to log in to the server I have set up in my office. Next, VS Code will prompt you to enter your password, which is your Lawrence ID number.
The first time you log into your account on one of the servers you will have to install both the Python extension and the Jupyter extension on the remote server. To do this, click the Extensions tab in VS Code and search for the Python and Jupyter extensions. VS Code will give you the option to install each extension on the server you are connected to.
Once you are connected and the extensions have been installed you can open a notebook to work on. The simplest way to do this is to select the Open... command from the File menu and then click the Show Local button. This will switch you to a dialog box that will allow you to select a file from your laptop to open.
Once you have the notebook open you should click the Select Kernel button in the top right corner of the notebook to select a kernel. The servers only have one copy of Python installed at /bin/python.
Finally, when you are done working you should save your work and then click on the >< rectangle in the bottom left corner of the window and select the Close Remote Connection command.
By now you should have downloaded the archive containing the author's notebooks. Try running some of those notebooks in this system to confirm that everything is working as it should.
You should read chapters 2-4 in the textbook and review the examples from those chapters. You only need to review the examples that show how to use the high level keras API - you can safely skip over the other examples where the author talks about lower level implementation details.