Tensorflow/Keras-tutorial | TensorFlow/Keras 教程

introduction

I notice that TensorFlow doesn't offer Chinese version tutorial/learning material in part of chapters for Chinese users, and the TensorFlow community '[email protected] Google Group' that provides Chinese translate service is unavailable now. So, I decide to create a project to translate the TensorFlow/Keras tutorial for Chinese users.

我注意到TensorFlow的一些教程和学习资料并没有为中国用户提供中文版本，且无法联系上提供翻译内容的TensorFlow社区'[email protected] Google Group'。所以我决定创建一个项目来为中文用户翻译TensorFlow/Keras的教程内容。

The features of this project are:

1. offer TensorFlow Keras officical tutorial guide Chinese version
2. add my personal comments/notes in each chapter

这个项目的特点有：

1. 提供了TensorFlow Keras官方指导的中文版本
2. 在每一个章节加入了我个人的评价和笔记内容

Note:

1. all code and English content are from TensorFlow Keras officical tutorial guide
2. the Chinese version will contain many errors due to my English level and ML understanding
3. TensorFlow Keras officially provides the Chinese version of some contents, you can switch the language setting in the guide page to read Chinese version.
4. If the Chineses version of a chapter is provided, I will note it in the catalog part and stop translate them.
5. For chapter without official Chinese version, I will offer the translated version link and the original link; for chapter with official chinese version, if I have finished the translated work, I will offer the translated version link. If I have not finished (and stop to translate in the future), I will offer the official Chinese version link.

请注意:

1. 所有的代码和英文内容来源于TensorFlow Keras官方指导版本
2. 囿于本人的英文水平和机器学习知识储备，中文版本不可避免的会含有一些错误，欢迎指出
3. TensorFlow Keras已经提供了部分内容的官方中文版本，你可以在对应的指导页面通过切换语言的方式来查看官中版本.
4. 如果某章节已经存在官中版本，我会在目录中注明并不再翻译这些内容.
5. 对于没有官中版本的内容，我会提供翻译版和原版的链接。对于有官中版本的内容，如果我已经完成翻译工作，我会提供翻译版的链接。如果我还没有完成翻译工作（之后也不会在翻译了），我会提供官中版本的链接（官中版本和原版是相同的联结，显示的语言由页面的语言设定所决定）

Catalog

1. The Sequential model
2. The Functional API (official Chinese version released)
3. Training and evaluation with the built-in methods (official Chinese version released)
4. Making new Layers and Models via subclassing (official Chinese version released)
5. Save and load Keras models (official Chinese version released)
6. Working with preprocessing layers
7. Customize what happens in Model.fit (official Chinese version released)
8. Writing a training loop from scratch (official Chinese version released)
9. Recurrent Neural Networks (RNN) with Keras
10. Masking and padding with Keras (official Chinese version released)
11. Writing your own callbacks (official Chinese version released)
12. Transfer learning and fine-tuning (official Chinese version released)
13. Training Keras models with TensorFlow Cloud

目录

1. 顺序模型 (翻译版，原版)
2. 函数式API (翻译版，原版/官中版)
3. 使用内置方法进行训练和评估 (翻译版, 原版/官中版)
4. 通过子类化创建新的层和模型 (翻译版，原版/官中版)
5. 保存和加载 Keras 模型 (翻译版，原版/官中版)
6. 使用预处理层 (翻译版，原版)
7. 自定义 Model.fit 的内容 (原版/官中版)
8. 从头编写训练循环 (原版/官中版)
9. 采用Keras的循环神经网络(翻译版，原版)
10. Keras 中的遮盖和填充(原版/官中版)
11. 编写自己的回调函数 (原版/官中版)
12. 迁移学习和微调 (原版/官中版)
13. 使用TensorFlow Cloud来训练Keras模型(翻译版，原版)