Skip to content

Recitation 8: Large Language Models

Overview

During this recitation, you will be able to run and evaluate a large language model on an example question/answer dataset. This is an individual activity so everyone should be running the notebook; however, feel free to discuss the activity and your approach with other students!

Instructions

For this recitation and Project 4, you'll be using the OpenAI Platform in a Colab notebook to query a large language model on a question/answer dataset. First, we need to set up our Azure project to enable the API and create a copy of the Colab notebook. Note: if you have not yet set up Azure, please refer to the instructions from Recitation 7.

  1. Open the Colab notebook using this link. Click on File --> Save as a copy in Drive to create your own copy that you will work on.

  2. Click on Share and make sure the Colab notebook is editable by anyone in Carnegie Mellon University. You will submit a link to this notebook for Gradescope.

  3. Follow the instructions in the notebook to set up Azure for LLM use.

  4. Have fun with the activity!

Once you are done, submit the link to your Colab notebook and the answers to the two questions in Gradescope at this link.