Retrieval Augmented Generation

In this project you are expected to apply Retrieval Augmented Generation to your project on antidepressants.

Assigned Reading

• Introduction to retrieval augmented generation (RAG) Read

• Relevant blogs: RAG optimization Web► Langchain Implementation Web►
• Relevant Python packages: Langchain Code► llamaindex Code►
• Generative AI Exists Because of the Transformer Blog►
• Language Models: A Guide for the Perplexed Read►

Assignments

Instruction for Submission of Assignments: Assignments should be submitted directly on Blackboard.  In rare situations assignments can be sent directly by email to the instructor. Submission should follow these rules:

1. Include a terse summary (not more than 1 sentence for each question) at start of your submission or in a separate file.  In the summary indicate if your responses agrees with responses posted to the open web site.
2. Submit your answers in a Jupyter Notebook Download► YouTube► Slides►
3. Submit your answers in Blackboard.

Task 1: Using PubMed, identify 10 studies that address effectiveness of antidepressants.  Create summary medication recommendations based on findings of the study using the text within the studies.

Task 2: Use Python to generate text from 500 studies on effectiveness of antidepressants. Include a reference to the original article in these auto-generated summaries.