What courses and learning materials does The Full Stack offer?

The Full Stack (fullstackdeeplearning.com) offers three main tiers of learning material: a 2023 Large Language Models Bootcamp focused on LLM-powered apps, a 2022 Deep Learning Course covering the full ML product lifecycle, and older FSDL course archives from 2018–2021. The site states its materials are shared for free, and the content spans lectures, labs, and project walkthroughs. Which tier fits you depends on whether you're building LLM applications, working on broader ML systems, or looking for foundational deep learning material.

2023 Large Language Models Bootcamp

This is the most current offering and the one to start with if your goal is building LLM-powered applications. The bootcamp covers the foundations of LLM apps, and its published sessions include:

• Launch an LLM App in One Hour — a fast, practical entry point
• LLM Foundations — background on how these models work
• Learn to Spell: Prompt Engineering — prompting as a skill
• Augmented Language Models — extending LLMs with external tools and data
• Project Walkthrough: askFSDL — a full example build
• UX for Language User Interfaces — designing the user-facing layer
• LLMOps — operating LLM apps in production
• What's Next? — closing perspective on the field

There are also guest sessions: Reza Shabani on training your own LLM, Harrison Chase on agents, and a fireside chat with Peter Welinder.

Choose this if: you want to ship an LLM application and need both the engineering and UX sides, not just model theory.

2022 Deep Learning Course (FSDL 2022)

This course treats an AI product as a full lifecycle problem, not a modeling exercise. Its lecture and lab pairs walk through:

There's also a project showcase and a course announcement.

Choose this if: you're building a broader ML-powered product and need to understand infrastructure, data, deployment, monitoring, and team process — the parts that sit around the model.

Older FSDL archives (2018–2021)

The site keeps earlier cohorts available, which is useful if you want fundamentals or a sense of how the curriculum evolved.

• FSDL 2021 — a synchronous online course with a full lecture/lab sequence: DL fundamentals, CNNs, RNNs, transformers, ML projects, MLOps infrastructure, troubleshooting deep neural networks, data management, AI ethics, testing & explainability, deployment & monitoring, research directions, and ML teams/startups. Includes a panel discussion, "Do I need a PhD to work in ML?", plus a course projects showcase.
• FSDL 2021 (Berkeley), FSDL 2020 (UW), FSDL 2019 (Online), FSDL 2019 (Bootcamp), and FSDL 2018 (Bootcamp) — earlier versions of the same program.

Choose this if: you want deep learning fundamentals (CNNs, RNNs, transformers) or you're working through the material in order and want the historical foundation the later courses build on.

How to decide

• Building an LLM app now → start with the 2023 LLM Bootcamp, especially "Launch an LLM App in One Hour" and the askFSDL walkthrough.
• Building a general ML product → the 2022 Deep Learning Course, since it covers the lifecycle from problem definition through monitoring.
• Learning fundamentals or filling gaps → the 2021 archive for CNNs, RNNs, and transformers before moving to the newer material.
• Just exploring → the site's blog and news sections sit alongside the courses, and the Multiplayer AI Manifesto is a short read on the site's philosophy of building AI products.

One caveat worth noting: the site's pricing signals point to a "subscribe" action, and the page states the course is shared "for free," but the material does not specify whether every resource or future offering carries the same terms. Treat the free framing as applying to the course materials described here rather than as a blanket statement about all site features.