What is d2l-pt?

The D2L.ai project (specifically this Portuguese translation) is an open-source, interactive textbook that teaches deep learning. Rather than just giving you theory or just giving you code, it combines explanations, math, diagrams, and runnable code all in one place so you can learn by doing. The entire book is built using Jupyter notebooks, which are documents that let you mix written explanations with live code. This means when a concept is introduced, you can immediately see and run the code that puts it into practice. It covers the math and concepts behind deep learning, but the goal is always practical: to help you build enough technical skill to start solving real problems in machine learning. The project is a community effort, updated frequently by both its authors and outside contributors. This resource is aimed at people who want to become applied machine learning scientists or engineers. For example, a student taking a university course on AI, a startup founder wanting to understand how deep learning works under the hood, or a product manager looking to get hands-on with the technology their team uses would all find this useful. The material is already being used in university classes around the world, and the project highlights several research papers that built on its tools. It is designed for people who want more than a surface-level overview and need a solid starting point for real-world work. The project is notable for being completely free and openly licensed, meaning anyone can access it and the community can help improve it over time. The book supports multiple deep learning frameworks, so you are not locked into learning just one toolset. It is a practical, hands-on alternative to traditional textbooks, built specifically to be updated quickly as the field of AI evolves.