Course · Advanced
From model to system
Context, reasoning, retrieval-augmented generation, mixture of experts and where it runs: what turns a model into a reliable assistant.
5 lessons · 1 quiz · 25 min · completion badge
What you will be able to do
- Explain what the context window is and how to use it without losing information.
- Describe what step-by-step reasoning changes and when to ask for it.
- Explain the principle of retrieval-augmented generation and its points of failure.
- Describe a mixture of experts and what it changes in cost and speed.
- Choose where to run a model according to the sensitivity of the data.
- Who it is for
- AI champions, IT managers and anyone who designs assistants or chooses an architecture.
- Prerequisites
- The course “Inside a language model”.
- Duration
- 25 min
- Level
- Advanced
Programme
What the course covers.
- 01Working with context3 lessons
- The context window6 min
Everything the model can take into account at once, and why the middle of a long document is the most fragile part.
- Reasoning step by step3 min
Why asking for the intermediate steps improves answers to multi-step problems.
- Search before answering: retrieval-augmented generation3 min
The principle that lets a model answer from your documents and cite its sources.
- The context window6 min
- 02Architecture and execution2 lessons
- Mixture of experts3 min
How a model can have a great many parameters while using only some of them for each token.
- Where a model runs3 min
Online, on a workstation or on your own servers: what changes for your data, your costs and your uses.
- Mixture of experts3 min
- ✓Final quiz6 questions