Week 7 of 12

A real model, deployed for real

Train an image classifier with fast.ai on Colab and deploy it as a live public Gradio app on us-central1 Cloud Run (NOT HuggingFace Spaces).

Why this resource? 60 second framing

Open the curated resource above and come back here when you have run it or watched it once. The coach is the right place to talk about what you saw.

What you ship by Sunday

Public Cloud Run service hosting a Gradio image classifier the student trained (e.g., samosa vs kachori), packaged with a Dockerfile (python:3.11-slim base) that copies the exported fast.ai model file and runs a Gradio server on port 8080; deployed via gcloud run deploy --source . --region us-central1 --memory 1Gi --allow-unauthenticated --min-instances 0 from Cloud Shell. Linked from a public GitHub repo with the training notebook, Dockerfile, requirements.txt, and deploy commands. README has the live .run.app URL, one screenshot of a real prediction, and a paragraph on the cold-start tradeoff (min-instances 0 means the first request after idle takes 5 to 15 seconds to wake the container; this is the price of zero cost).

The coach (Socratic, one hint per turn)

30 messages / day

The coach gives you one hint per turn. Never the answer. Hinglish bilkul welcome hai. It cannot do arithmetic; run code yourself and paste the result.

Ask for a Socratic review of your artifact

Paste your README (or a draft of it). The coach will read it like an honest senior would, ask one question that surfaces the weakest link, and point at one thing to tighten. It will not rewrite it for you. Maximum 8000 characters.

Before you close the tab

Three short questions to answer in your own words, even just in your head. The point is to notice what you actually learned this week, not to perform learning back at the coach.

What broke first when you tried this week artifact, and what did you do?
Pick one line of code that confused you initially. What does it do?
If you had to rebuild this from a blank file tomorrow, what would you start with?