This repository provides a comprehensive tutorial for deep learning, based on Professor Hung-yi Lee's popular university course. It aims to make complex deep learning concepts accessible to beginners through clear explanations, detailed derivations, and engaging examples, catering to students and practitioners seeking a structured Chinese-language introduction to the field.

How It Works

The tutorial distills and optimizes content from Professor Lee's "Machine Learning" course, offering detailed derivations for formulas and in-depth explanations of challenging topics. It also incorporates material from other courses and supplementary deep learning knowledge to provide a well-rounded understanding.

Quick Start & Requirements

• PDF Download: Available at https://github.com/datawhalechina/leedl-tutorial/releases and a domestic Baidu Netdisk link.
• Companion Code: Located in the Homework folder within the repository.
• Errata: Available at https://datawhalechina.github.io/leedl-tutorial/#/errata.

Highlighted Details

• Covers a wide range of deep learning topics including CNNs, Transformers, GANs, Self-Supervised Learning (BERT, GPT-3), Autoencoders, Diffusion Models, Explainable AI, Transfer Learning, Deep Reinforcement Learning, Meta-Learning, and ChatGPT.
• Includes detailed mathematical derivations for key concepts.
• Offers a structured curriculum with content sourced from Professor Hung-yi Lee's renowned lectures.
• Provides companion code for practical implementation and learning.

Maintenance & Community

• Contributors include Qi Wang (SJTU), Yiyuan Yang (Oxford), and Ji Jiang (Peking University).
• A WeChat official account (Datawhale) is available for community engagement and discussion groups.

Licensing & Compatibility

• Licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0).
• This license restricts commercial use and requires derivative works to be shared under the same terms.

Limitations & Caveats

The tutorial is primarily a textual resource with accompanying code; it does not appear to be an interactive platform or a software library for direct integration into other projects. The CC BY-NC-SA 4.0 license may limit its use in commercial or closed-source applications.