Every module includes three substantial lessons, downloadable source files, practical exercises, explained solutions and a knowledge check. Modules marked free are available under the current public access rules. Protected self-learning modules are free after registration and email verification. Publication, prerequisites and explicit private-access rules still apply.
01Free module
Neural Networks and Backpropagation
Connect predictions, loss and gradients through an inspectable numerical example.
- A neuron/loss calculation
- An analytic/finite-difference gradient comparison
- An optimization and evaluation plan
Overview & practice files ↗View lesson outline
- Trace a neuron and its loss
- Derive and check a gradient
- Separate optimization from generalisation
02Free module
PyTorch Fundamentals
Use a small explicit CPU training loop and preserve evaluation boundaries.
- An annotated tensor/split map
- An executed CPU training loop
- A frozen validation and test-boundary note
Overview & practice files ↗View lesson outline
- Represent data as tensors with explicit shape
- Implement forward, backward and optimizer steps
- Evaluate without changing the model
03Free · verified account
Computer Vision
Trace image tensors, convolution shapes and validation on generated stroke images.
- An image/axis specification
- A checked convolution shape calculation
- An executed toy validation and sensitivity plan
View access requirements ↗View lesson outline
- Define image representation and safe scope · Verified account
- Trace a small convolutional model · Verified account
- Evaluate image errors and dataset limits · Verified account
04Free · verified account
Natural Language Processing
Build a train-vocabulary text classifier and examine unknown tokens and split discipline.
- A label/tokenization contract
- An executed count-vector classifier
- An error, unknown-token and privacy review
View access requirements ↗View lesson outline
- Define text labels and preprocessing · Verified account
- Train a simple numerical text representation · Verified account
- Inspect errors and protect text privacy · Verified account
05Free · verified account
Large Language Models, Embeddings, and Transformers
Understand representations and attention while separating a toy calculation from a trained LLM.
- A representation comparison
- A checked attention calculation
- A grounded-answer and adversarial evaluation plan
View access requirements ↗View lesson outline
- Compare tokens, sparse vectors and embeddings · Verified account
- Calculate scaled dot-product attention · Verified account
- Evaluate context, hallucination and retrieval limits · Verified account
06Free · verified account
Practical Generative AI Application
Build a documented document-search baseline with citations, abstention and disabled API integration.
- A retrieval/application contract
- An executed citation/abstention baseline
- An evaluation and disabled-integration handover
View access requirements ↗View lesson outline
- Define a source-grounded assistant contract · Verified account
- Implement retrieval and inspect its limits · Verified account
- Write an evaluation, privacy and cost handover · Verified account
07Free · verified account
Controlled agents, tool calling and document workflows
Implement a bounded tool-calling contract: use an original fictional source to implement, verify and explain three connected professional assignments.
- Source/metric/access contract
- Independent implementation and exact run instructions
- Unchanged control and changed-input/failure evidence
- Decision memo with limitations and next evidence
View access requirements ↗View lesson outline
- Implement a bounded tool-calling contract · Verified account
- Inspect a vector store and retrieval boundary · Verified account
- Classify documents and review unsupported claims · Verified account
