CS 163 — Web Apps, Cloud Deployment & ML Fine-Tuning

Course materials covering interactive web applications, cloud deployment, containerization, and model fine-tuning.

Repository Structure

Directory	Topic
dashapps/	Plotly Dash web apps (HTML layout, data tables, graphs, callbacks, multi-page apps)
appengine/	Deploying a Dash app to Google App Engine with Cloud Storage integration
intro-to-docker/	Containerizing a FastAPI ML inference service with Docker and deploying to Cloud Run
fine-tuning/	Transformer fine-tuning techniques: linear probing and LoRA

Topics Covered

Dash Web Apps (dashapps/)

Progressive examples building up from basic HTML to interactive dashboards:

• app1.py – basic layout
• app2.py – HTML and styling
• app3.py – data tables and charts
• app4.py – callbacks and interactivity
• app5-multi/ – multi-page app
• app6.py – hover/click graph interactions
• app7.py / app8.py – Bootstrap themes and layout

Google App Engine (appengine/)

• Deploying a Dash app with Gunicorn via app.yaml
• Reading data from Google Cloud Storage using google-cloud-storage
• Environment variables for bucket configuration

Docker & Cloud Run (intro-to-docker/)

• FastAPI REST API serving a Decision Tree model
• Writing a Dockerfile and building images
• Publishing to Google Artifact Registry
• Deploying to Cloud Run

Fine-Tuning (fine-tuning/)

• 01_probing.ipynb – Linear probing on DistilBERT for IMDb sentiment classification
• 02_lora.ipynb – Parameter-efficient fine-tuning with LoRA

Notes

See note.md for class notes and references on Dash and Google App Engine topics.