Module 6: Transformers
1 min read·13 Oct 2026
The architecture behind every model this course calls, its three families, how a model is trained and adapted, when to adapt at all, and what serving one costs. Every lesson runs a small open model on your CPU.
What you need first
- Module 0 (calling a model with the standard setup cell)
- Module 3 (train and test sets)
- Module 4 (embeddings; the training loop)
- No GPU needed