Lecture 14

Agenda

• Recap of what we learned on Tuesday
• Discussion of Chrome extensions
• Filling out FINAL PROJECT GROUP SURVEY
• Ideation discussion for final project
• (If time) Local LLM Integration Project

What are we learning? (1)

• Client/server architectures
  • With Render you deploy a server
  • With local code you are creating a client
• "Standardization"
  • We have decided on a "standard"
  • Send:
    • { "password": "<password to check>" }
  • Respond:
    • { "valid": true/false, "reason": "<human readable error reason>" }
  • Importantly, the "reason" is not software readable (it would require standardizing the error strings)
• This allows many servers to interact with many clients

What are we learning? (2)

• Learning how to deploy from GitHub
• Learning about "Infrastructure as Code" (render.yaml)
• Learning about
  • remote deployment (https://cis3500-buggy-password-validator.onrender.com/)
  • local deployment (http://localhost:5324)
• Learning how to make GET/POST requests
  • with CURL in the terminal
  • with requests.get() and requests.post() in Python
  • with RapidAPI or Postman
• Learning about JSON, content-types, parsing

What are we learning? (3)

• "Grokking"
  • "To understand something so deeply and holistically that it becomes second nature."
  • (Martian word from sci-fi author Robert A. Heinlein in 1961 novel Stranger in a Strange Land)
• The kinds of problems you may encounter
  • Uses Python 3 but you don't have it
  • Uses Python 3.12 but you have Python 3.11
  • Documentation shows endpoint that isn't yours
  • Forgot to install packages
  • Forgot to use virtual environment (or pipenv)

Chrome Extension

Presentation

Ideation

Data Extractors

Data Extractors

• Extract data from your context
• That is useful to paste in ChatGPT
• Create extensions that simplify this process

• Ideation workshop now
  • Activity as a group (15 mins)
    • Download the LLM Use Case Sheets
    • Create a list of all the data types that are necessary to create prompts
  • Activity as an individual (10 mins)
    • Come up with ideas of data extractors
    • Design a Chrome extension idea
  • Activity as group (10 mins)