2. Building Models🔗

These chapters start with the public surface: tensors, shapes, model builders, datasets, and small training loops. They are intentionally practical. They show how a TorchLean model is written and run before asking you to care about the IR or proof machinery underneath it.

1. 2.1. Tensors, Shapes, and DTypes
2. 2.2. Building Models With Layers
3. 2.3. Datasets, Loaders, and Minibatches
4. 2.4. Training From Scratch
5. 2.5. TorchLean API