21 TensorFlow projects from the Python 2 era
Companion code for a Chinese hands-on deep learning book covering CNNs, GANs, RNNs, and reinforcement learning chapter by chapter.

What it does
This repository holds the runnable companion code for the Chinese book 21 Projects to Play with Deep Learning — Detailed Practice Based on TensorFlow. It walks through 21 chapter-sized projects spanning CNNs (including GAN variants like DCGAN, pix2pix, and CycleGAN), RNNs and LSTMs, and reinforcement learning from Q-Learning to Policy Gradient. Each chapter maps to a self-contained folder with source code, while separate Baidu Pan downloads host the larger datasets.
The interesting bit
The value is in the breadth and structure rather than novelty: it treats deep learning as a craft best learned by building Deep Dream, style transfer, neural machine translation, and image captioning in sequence. For readers of the book, it provides a curated, end-to-end tour of what constituted state-of-the-art hobbyist deep learning circa 2016–2018.
Key highlights
- 21 projects covering three arcs: computer vision (chapters 1–11), sequence modeling (12–17), and reinforcement learning (18–21)
- Includes era-specific favorites: Deep Dream, super-resolution, Char RNN, word2vec, DQN, and SARSA
- ~4,700 GitHub stars, suggesting a sizable Chinese-language readership
- Supplementary data files hosted on Baidu Pan with extraction code
1kmf
Caveats
- Locked to a legacy stack: the README recommends Python 2.7, TensorFlow ≥1.4.0, and Ubuntu 14.04
- Entirely Chinese-language; no English translation of documentation or comments
- Data is hosted on Baidu Pan, which may be inaccessible or inconvenient outside mainland China
Verdict
A useful reference if you read Chinese and want a historical, project-driven survey of TensorFlow 1.x techniques. Everyone else should probably reach for modern PyTorch or TensorFlow 2 tutorials instead.
Frequently asked
- What is hzy46/Deep-Learning-21-Examples?
- Companion code for a Chinese hands-on deep learning book covering CNNs, GANs, RNNs, and reinforcement learning chapter by chapter.
- Is Deep-Learning-21-Examples open source?
- Yes — hzy46/Deep-Learning-21-Examples is an open-source project tracked on heatdrop.
- What language is Deep-Learning-21-Examples written in?
- hzy46/Deep-Learning-21-Examples is primarily written in Python.
- How popular is Deep-Learning-21-Examples?
- hzy46/Deep-Learning-21-Examples has 4.7k stars on GitHub.
- Where can I find Deep-Learning-21-Examples?
- hzy46/Deep-Learning-21-Examples is on GitHub at https://github.com/hzy46/Deep-Learning-21-Examples.