Jupyter Setup Guide v.2 - cs327e-summer2022/snippets GitHub Wiki

Follow this guide to set up your Jupyter notebook environment for this class.

• Go to the GCP console.
• Type "Vertex AI" in the search bar and go to the Vertex AI page.
• Click on Workbench from the left-hand menu.
• You should see an Enable Notebooks API link at the top of the page. Click on it to Enable the Notebooks API.
• Once the API is enabled, click on the + New Notebook button to create a new instance.
• Choose Python 3.
• On the new notebook screen, enter a name for your notebook (e.g. notebook).
• For the region, choose us-central1 (Iowa).
• For the zone, accept the default value.
• Click the Advanced Options link to customize your notebook.
• In the environment section, choose Debian 10 and Python 3 with Intel MKL.
• Choose n1-standard-8 for the machine type.
• Do not change the disk settings. They should both be Standard Persistent Disk and sized at 100G for the boot disk and 100G for the data disk.
• Don't make any other changes.
• Click the Create button at the bottom of the page to create your Jupyter instance.
• From the top left nav menu, go to IAM and find your Compute Engine default service account. Click on the pencil icon next to it and add the role "Storage Object Admin" under the "Cloud Storage" group. Do not remove other roles. Save the changes.
• This completes your notebook setup. Please remember to stop your notebook when you done using it. If you forget to stop it, you will quickly run out of GCP credits! You can stop the notebook by selecting the notebook and clicking on Stop. Try to get in the habit of stopping your resources whenever you are not using them.