12.8K ·GitHub
What is this?
This is the companion code repository for the Chinese book 'Deep Learning Framework PyTorch: Introduction and Practice', covering PyTorch basics and fun projects like neural talk, poem writing, anime face generation, and style transfer.
This is the companion code repository for the Chinese book 'Deep Learning Framework PyTorch: Introduction and Practice', covering PyTorch basics and fun projects like neural talk, poem writing, anime face generation, and style transfer.Is this a standalone tutorial or does it require the book?You do not need to buy the book: the repo contains over 60% of the book's text and over 90% of its code, especially the early introductory chapters are nearly fully covered.
You do not need to buy the book: the repo contains over 60% of the book's text and over 90% of its code, especially the early introductory chapters are nearly fully covered.How do I get started with this?Install PyTorch following the official instructions for your platform, then clone the repository and open the notebooks.
Install PyTorch following the official instructions for your platform, then clone the repository and open the notebooks.Which PyTorch version does the code target?The code was mainly tested with Python 3 and PyTorch 1.6 to 1.8; Python 2 and newer PyTorch versions are untested.
The code was mainly tested with Python 3 and PyTorch 1.6 to 1.8; Python 2 and newer PyTorch versions are untested.Are the notebook explanations fully polished?Some of the markdown text is an early draft of the book and about 80% consistent with it, so there may be awkward phrasing or minor errors that the author is gradually fixing.
Some of the markdown text is an early draft of the book and about 80% consistent with it, so there may be awkward phrasing or minor errors that the author is gradually fixing.How is the content structured?It is organized in three parts: basic PyTorch usage (Chapters 2-5: Tensor, autograd, nn, data/visualization tools), advanced topics (Chapters 6-8: vectorization, distributed training, CUDA extensions), and practical applications (Chapters 9-13: image classification, GAN anime face generation, poem writing with Transformer/CharRNN, style transfer, and CenterNet object detection).
It is organized in three parts: basic PyTorch usage (Chapters 2-5: Tensor, autograd, nn, data/visualization tools), advanced topics (Chapters 6-8: vectorization, distributed training, CUDA extensions), and practical applications (Chapters 9-13: image classification, GAN anime face generation, poem writing with Transformer/CharRNN, style transfer, and CenterNet object detection).Can I write poems or generate anime faces with this?Yes, Chapter 11 implements a Transformer/CharRNN-based model that writes classical Chinese poetry (including acrostic poems), and Chapter 10 builds a GAN from scratch that generates anime avatar faces in varying styles.
Yes, Chapter 11 implements a Transformer/CharRNN-based model that writes classical Chinese poetry (including acrostic poems), and Chapter 10 builds a GAN from scratch that generates anime avatar faces in varying styles.Are pretrained models provided for the demo projects?Yes, for the practical application chapters the repo provides complete implementation code along with pretrained models so readers can test the demos directly.
Yes, for the practical application chapters the repo provides complete implementation code along with pretrained models so readers can test the demos directly.Is the material in Chinese or English?The notebooks and accompanying text are in Chinese, matching the Chinese-language book they are based on.
The notebooks and accompanying text are in Chinese, matching the Chinese-language book they are based on.What are similar projects?DeepLearning, pytorch-handbook, TensorFlow-Examples, deeplearning-models, fastbook, autumn-recruitment, PyTorch-GAN, CV
DeepLearning, pytorch-handbook, TensorFlow-Examples, deeplearning-models, fastbook, autumn-recruitment, PyTorch-GAN, CV