Dive-into-DL-TensorFlow2.0

Dive-into-DL-TensorFlow2.0 Project Overview

Introduction

Dive-into-DL-TensorFlow2.0 is an innovative project that aims to transform the code examples from the well-known book Dive into Deep Learning, originally implemented in MXNet, into implementations using TensorFlow 2.0. The project has received approval from the original author Mu Li and is primarily based on the Chinese edition of the book. Additionally, Dive-into-DL-TensorFlow2.0 takes inspiration from a similar project, Dive-into-DL-PyTorch, focusing on converting the book’s content into PyTorch.

Project Details

Hosted on GitHub, Dive-into-DL-TensorFlow2.0 is structured into two main folders: code and docs. The code folder includes Jupyter notebook files for each chapter, showcasing TensorFlow 2 implementations. The docs folder contains markdown documentation of the book's content, which is published as a static website using docsify. This way, the documentation can be accessed easily with some differences from the original work due to the transition from MXNet to TensorFlow.

Target Audience

This project is particularly beneficial for those interested in learning about deep learning, specifically through TensorFlow 2. Dive-into-DL-TensorFlow2.0 is suited for learners with a basic understanding of mathematics and programming, such as linear algebra, calculus, probability, and Python programming, but does not require prior deep learning experience.

How to Use the Project

Method 1: Access Online

The easiest way to explore the documentation with mathematical formulae is by accessing the web version directly available via this link. However, to run code examples locally, users need to clone the repository and execute them from the code directory.

Method 2: Local Access

For offline access, you can set up the project locally. First, install the docsify-cli tool using the following command:

npm i docsify-cli -g

Then, clone the repository:

git clone https://github.com/TrickyGo/Dive-into-DL-TensorFlow2.0
cd Dive-into-DL-TensorFlow2.0

Finally, run a local server to view the rendered documents at http://localhost:3000:

docsify serve docs

Contributors

The project has benefited from the contributions of several key individuals:

• archersama - Project leader and core contributor.
• TrickyGo - Contributor.
• SwordFaith - Contributor.
• ShusenTang - Contributor.
• LIANGQINGYUAN - Contributor.

Table of Contents

The project covers a broad range of topics from the book:

1. Introduction to Deep Learning
2. Preliminaries - Installation, data manipulation, and more.
3. Deep Learning Basics - Linear regression, softmax regression, and MLPs.
4. Deep Learning Computation - Model construction, parameter management, and more.
5. Convolutional Neural Networks - Covering layers, structures, and advanced networks.
6. Recurrent Neural Networks - Language models, GRUs, LSTMs, and more.
7. Optimization Algorithms - Techniques including gradient descent variants.
8. Computational Performance - Performance-improving methods with programming strategies.
9. Computer Vision - Image processing, segmentation, and more.
10. Natural Language Processing - Word embeddings, sentiment analysis, and seq2seq models.
11. Appendices - Useful supplementary materials and guides.

The content is continuously updated to include the latest developments in the field.

Original Book

• Chinese Version: 动手学深度学习 | GitHub Repository
• English Version: Dive into Deep Learning | GitHub Repository

Citation

For academic use, please cite the original book in your research:

@book{zhang2019dive,
    title={Dive into Deep Learning},
    author={Aston Zhang and Zachary C. Lipton and Mu Li and Alexander J. Smola},
    note={\url{http://www.d2l.ai}},
    year={2019}
}

This project is a valuable resource for anyone looking to delve into the world of deep learning using TensorFlow 2.0, offering comprehensive insights from the basic concepts to advanced neural network architectures and applications.