16  Neural Networks, Deep Learning, and Transformers: torch (todo)

Note

Proposed chapter scope: Using pre-trained neural networks and transformer models via torch. Emphasis on using models rather than building them from scratch — fine-tuning a pre-trained model for a classification task is the target level of ambition. Provides the conceptual grounding for what tidyllm (chapter 11) was doing under the hood and a better understanding of modern AI