What it does This repo is a hand-picked reading list of deep-learning resources—playgrounds, courses, tools, papers, and blog posts—organized from beginner to advanced. The author, a Taiwan-based developer, maintains it as a living notebook of things he actually used while learning.

The interesting bit Most “awesome” lists are flat Markdown dumps. This one is structured like a curriculum: browser demos first (TensorFlow Playground, GAN Lab, Quick Draw), then courses (Andrew Ng, fast.ai, 李宏毅), then tooling and papers. It is also unapologetically Traditional Chinese, which makes it a rare on-ramp for Chinese-speaking developers who find English-only lists intimidating.

Key highlights

• Playground section is the star: ~15 interactive demos (ConvNetJS, Teachable Machine, NVIDIA GauGAN, even a Waifu Vending Machine) that let you see gradients flow before you read a single equation.
• Courses cover the usual suspects—Coursera Deep Learning Specialization, fast.ai, CS231n, CS224n—plus Taiwan-specific picks like 李宏毅’s NTU lectures.
• Tools include practical favorites: Colab, TensorBoard, Papers With Code, BERTViz, the What-If Tool.
• Papers and blogs are annotated with links to the author’s own blog posts when he has written walkthroughs (e.g., Transformer and GPT-2 tutorials).
• Explicitly marked as “continuously updated,” though the README does not state a cadence.

Caveats

• The list is heavy on links and light on commentary; do not expect original code or exercises.
• Some sections (tutorials, papers) are sparsely annotated compared to the playground and courses.
• A few demo links may rot; the author notes “to be updated” placeholders in some sections.

Verdict Great if you are a Traditional Chinese speaker looking for a structured, opinionated path into deep learning. Skip it if you already have your own curated pipeline or need hands-on projects rather than pointers.