A Chinese-language PyTorch curriculum that runs in your browser

What it does This repository is a Chinese-language adaptation of GokuMohandas’s practicalAI, structured as a sequence of Jupyter notebooks that teach machine learning and deep learning with PyTorch. The curriculum starts with Python, NumPy, and Pandas, moves through classical algorithms such as linear regression and random forests, and advances to CNNs, RNNs, GANs, and reinforcement learning. Every notebook is intended to run directly in Google Colab, removing the need for local environment setup.

The interesting bit The authors claim the material is “not just tutorials” but an introduction to product-level, object-oriented machine learning programming—a stated goal that sets it apart from typical quick-script notebook courses. Whether every notebook lives up to that engineering standard is unclear from the README alone, but the explicit ambition is notable.

Key highlights

• Covers a broad spectrum, from KMeans clustering and random forests to spatial transformer networks and pretrained language modeling.
• Localized for Chinese-speaking developers; several notebooks credit individual translators, suggesting a community-driven effort.
• Runs entirely in-browser via Google Colab, requiring no local GPU or dependency management.
• Uses PyTorch consistently across all deep-learning and advanced sections.

Caveats

• The translation appears partially incomplete: the contribution table lists no translator for several notebooks, including Convolutional Neural Networks, Recurrent Neural Networks, and Computer Vision.
• The README offers no detail on how the promised “product-level OOP” approach is actually implemented in the notebook code.

Verdict Useful if you read Chinese and want a broad, zero-install survey of modern ML. If you are comfortable with the original English practicalAI, this fork adds language accessibility but not new content.