AI201 Lab 3 — Pod Classifier

A few-shot podcast episode classifier. Given an episode description, classifies it as interview, solo, panel, or narrative using labeled examples and an LLM.

Setup

python -m venv .venv
source .venv/bin/activate     # Mac/Linux
# or: .venv\Scripts\activate  # Windows

pip install -r requirements.txt
cp .env.example .env
# Add your GROQ_API_KEY to .env

Run

python app.py

Lab milestones

Milestone	Task	File
1	Label 20 training episodes	data/my_labels.json
2	Implement the few-shot classifier	classifier.py
3	Implement evaluation metrics	evaluate.py

See the lab instructions for full details.

Project structure

ai201-lab3-podclassifier-starter/
├── app.py              # Gradio UI
├── classifier.py       # Few-shot classification logic
├── evaluate.py         # Evaluation metrics
├── config.py           # Settings and constants
├── requirements.txt
├── .env.example
├── data/
│   ├── train_episodes.json   # 20 episodes to label (Milestone 1)
│   ├── test_episodes.json    # 20 pre-labeled episodes (held-out test set)
│   ├── my_labels.json        # Your labels — edit this in Milestone 1
│   └── taxonomy.md           # Label definitions and edge cases
└── specs/
    ├── system-design.md      # Architecture overview
    ├── classifier-spec.md    # Spec for Milestone 2 (fill in before coding)
    └── evaluation-spec.md    # Spec for Milestone 3 (fill in before coding)