Open-source deep learning textbook with a Chinese edition
Introduction
Dive into Deep Learning (D2L), authored by Mu Li and colleagues, is the open-source textbook pairing rigorous derivations with runnable code, adopted by 500+ universities worldwide. The Chinese edition is excellent, with companion Bilibili videos — the first-choice textbook for systematic DL study in Chinese.
Key Features
- Theory plus runnable code
- Adopted by 500+ universities
- High-quality free Chinese edition
- Companion Bilibili video course
- MXNet/PyTorch/TensorFlow coverage
Pros
- The ceiling of Chinese DL textbooks
- Completely free and open
Cons
- Dense content demands patience
Pricing
Completely free and open-source.（See official site for current pricing）
FAQ
Prerequisites?
Linear algebra, calculus and Python basics.
Which framework version?
The PyTorch edition is most popular.
Conflicts with Ng courses?
Complementary — videos build intuition, the book cements details.
Related Sites
Collected on lucifer.city · Reviewed by the founder