AI
Attention in Transformers: Concepts and Code in PyTorch
Train your own R1 reasoning model with Unsloth (GRPO) Llama3.1_(8B)-GRPO.ipynb
LLMs-Zero-to-Hero，完全从零手写大模型，从数据处理到模型训练
AI 大神免費深入淺出全面講解大型語言模型、訓練、心理學到實際應用，一次搞定！OpenAI 共同創辦人、特斯拉人工智慧總監 Andrej Karpathy
目前可用的 DeepSeek R1 模型 API 服务商(2025年2月)
Synthetic Data Generation with LLMs
The DeepSeek Series: A Technical Overview
Programming
21st Century C++ By Bjarne Stroustrup
7 Common Mistakes in Architecture Diagrams
Introduction to Domain-Driven Design
Other
硅谷视角深聊：DeepSeek的颠覆、冲击、争议和误解【硅谷101】
Bill Gates on Microsoft at 50, and what’s next for AI and innovation
Make McKinley Great Again -How Trump is bringing the 19th century into the 21st