Module 12: PyTorch

The deep learning toolchain for the rest of the course. This module covers PyTorch from the ground up — tensors and data pipelines, then the mechanics of building, training, and structuring real models.

What You’ll Learn¶

• Tensors: creation, manipulation, broadcasting, and GPU acceleration

• Dataset and DataLoader pipelines for batching and shuffling

• Building models with torch.nn.Module

• Automatic differentiation and the training loop

• Activation functions, loss functions, and optimizers

• Structuring larger projects and higher-level training frameworks

Lecture Materials¶

• 📊 Lecture Slides (PDF)

• 📓 Chapter notebooks: PyTorch fundamentals (two parts, ch12) and the mechanics of PyTorch (two parts, ch13)

Prerequisites¶

• Module 10 (neural network fundamentals, backpropagation)

Reference¶

Raschka & Mirjalili, Machine Learning with PyTorch and Scikit-Learn, Chapters 12–13.