Setup
Ten Minutes, Before Day 1
1. Open the Setup Notebook
All labs run in Google Colab, so there is nothing to install. You need a Google account.
Open the setup notebook in Colab and choose Runtime → Run all. It checks the runtime and tells you which provider the API labs will use. Every other notebook is on the notebooks page.
For the labs that need a GPU, switch one on: Runtime → Change runtime type → T4 GPU. Labs 1, 9, 11 and 12 run on a CPU. Every lab is designed for the free tier; the readiness page says which ones have been timed on it.
2. Add API Keys (optional)
Some labs call commercial APIs. Keys are optional: with none set, those labs run on a small open model from Hugging Face, and you can still complete every exercise.
If you have keys, add them to Colab Secrets (the key icon in the left sidebar), and switch on Notebook access for each:
| Secret name | Used in | Where to get a key |
|---|---|---|
OPENAI_API_KEY |
Modules 8, 11, 13–15 | platform.openai.com |
ANTHROPIC_API_KEY |
Modules 8, 11, 13–15 | console.anthropic.com |
TYPESAFE_API_KEY |
Modules 12–15 | typesafe.ai |
Never paste a key into a notebook cell. Colab Secrets keeps keys out of the notebook file, so they cannot be shared or committed by accident.
Each API lab states its approximate cost for one full run.
Install TypeSafe’s SDK only under its official name, typesafe-sdk; the labs install it for you. Several lookalike packages on PyPI, such as typesafe-ai and jev, are not published by TypeSafe.
3. Check the Prerequisites
Take the 15-minute entry check on the Before Day 1 page: it tells you, area by area, whether to review anything before Day 1 and which sections of which free resources to read.
4. Module 0: Optional Pre-Work
Module 0 · Coding Agents in the Terminal is optional pre-work: do it at home, before Day 1. If you get stuck, bring your laptop to the drop-in clinic from 08:00 to 09:00 on Day 1, before the welcome; it is optional too. Unlike the labs, Module 0 runs on your own laptop, not in Colab: you install a terminal coding agent, git and the GitHub CLI, so you need permission to install software and a GitHub account. Its page walks you through the installs. Nothing later in the workshop depends on it.
Running Locally (optional)
Colab is the supported environment. To run the notebooks on your own machine, install uv and run:
git clone https://github.com/project-delphi/nlp-llms
cd nlp-llms
uv run --group notebooks jupyter labSet the API keys as environment variables with the same names as the secrets above.