NLP to LLMs
  • Home
  • Start here
  • Schedule
  • Days
    • Before Day 1 · Module 0 (pre-work)
    • Day 1 · Foundations
    • Day 2 · Transformers and pretraining
    • Day 3 · Using and aligning LLMs
    • Day 4 · Calibration, decisions, RAG
    • Day 5 · Agents and capstone
  • Notebooks
  • References
  • Teach

Schedule

Five days, 09:00 to 17:00

On Day 1, each module is 95 minutes: 45 minutes of lecture, then 50 of lab. On Days 2 to 5, each module is 120 minutes: 55 minutes of lecture, 55 of lab, then a 10-minute debrief. On Days 2 to 4 the second module breaks for lunch between its lecture and its lab. On Day 5, the capstone build runs on across lunch. The capstone (Module 15) has no lecture: it takes 11:30–12:25, 13:25–14:30 and 14:45–16:45 on Day 5.

The intro slides, shown in the welcome slot on Day 1, give the setup and this schedule in short form.

Day 1 keeps the shorter modules because its lectures are the lightest. From Day 2, each lecture has room for its predictions, checks and demo, and each lab ends with a debrief: what the room found, and where results differed. Those days open with retrieval practice, a few exit-check questions from earlier days, and close with a synthesis: the day’s results go into a running table, and the room names the limitation that is left.

Module 0 · Coding agents in the terminal is optional pre-work, done on your own laptop before Day 1. There is no talk and no notebook. If you get stuck, the room opens at 08:00 on Day 1 for an optional drop-in clinic. The workshop proper starts at 09:00.

Lecture Lab Debrief Break Optional clinic

Day 1 · 95-minute modules

Time Day 1 Foundations
08:00–09:00 Drop-in clinic: Module 0 pre-work (optional)
09:00–09:10 Welcome, setup check
09:10–10:45 1 · Text as data Lecture 09:10Lab 09:55
10:45–11:00 Break
11:00–12:35 2 · Word vectors and neural networks Lecture 11:00Lab 11:45
12:35–13:35 Lunch
13:35–15:10 3 · Sequence models Lecture 13:35Lab 14:20
15:10–15:25 Break
15:25–17:00 4 · Seq2seq and attention Lecture 15:25Lab 16:10

Days 2–5 · 120-minute modules

Time Day 2 Transformers and pretraining Day 3 Using and aligning LLMs Day 4 Calibration, decisions, RAG Day 5 Agents and capstone
09:00–09:15 Retrieval practice Retrieval practice Retrieval practice Retrieval practice
09:15–11:15 5 · The transformer Lecture 09:15Lab 10:10Debrief 11:05 8 · LLMs through APIs Lecture 09:15Lab 10:10Debrief 11:05 11 · Calibration Lecture 09:15Lab 10:10Debrief 11:05 14 · Agents Lecture 09:15Lab 10:10Debrief 11:05
11:15–11:30 Break Break Break Break
11:30–12:25 6 · Pretraining and the Hugging Face stack Lecture 11:30 9 · Reinforcement and preference learning Lecture 11:30 12 · Calibrated decisions: RLCD and Jev Lecture 11:30 15 · Capstone: build Hands-on
12:25–13:25 Lunch Lunch Lunch Lunch
13:25–14:30 6 · Pretraining and the Hugging Face stack Lab 13:25Debrief 14:20 9 · Reinforcement and preference learning Lab 13:25Debrief 14:20 12 · Calibrated decisions: RLCD and Jev Lab 13:25Debrief 14:20 15 · Capstone: build Hands-on
14:30–14:45 Break Break Break Break
14:45–16:45 7 · Fine-tuning and LoRA Lecture 14:45Lab 15:40Debrief 16:35 10 · RLHF Lecture 14:45Lab 15:40Debrief 16:35 13 · Retrieval-augmented generation Lecture 14:45Lab 15:40Debrief 16:35 15 · Capstone: evaluate and share Hands-on
16:45–17:00 Synthesis Synthesis Synthesis Synthesis and wrap-up

From Traditional NLP to Modern LLMs · Genial Labs · Built with Quarto

  • Setup

  • FAQ

  • Readiness

Text CC BY 4.0 · Code MIT