The resource library
Code &
resources.
课程的另一面：示例代码、实验说明与可下载的讲义。按课程选择资源，与 Notes 中的网页讲义配合使用。
课程源码
保留原有目录结构 · 按课程下载 ZIP
配套代码用于学习与实验，尚未逐一运行验证。请先阅读课程页的说明，在独立环境中检查依赖后运行；包内不含模型权重和运行产物。
Stanford / CME295
Transformers and Large Language Models
Stanford / CS224N
Natural Language Processing with Deep Learning
Stanford / CS224V
Agentic AI
Stanford / CS229
Machine Learning
Stanford / CS230
Deep Learning
Stanford / CS234
Reinforcement Learning
Stanford / CS236
Deep Generative Models
Stanford / CS324
Advances in Foundation Models
Stanford / CS329A
Self-Improving AI Agents
Stanford / CS329H
Machine Learning from Human Preferences
Stanford / CS329T
Trustworthy Machine Learning
Stanford / CS329Z
Engineering AI Agents
Stanford / CS336
Language Modeling from Scratch
Stanford / CS349D
AI Inference Infrastructure
Stanford / CS349E
Efficient ML Inference at Scale
CMU / CMU17-630
Prompt Engineering
CMU / CMU10-423
Generative AI
CMU / CMU11-667
Large Language Models: Methods and Applications
CMU / CMU11-868
Large Language Model Systems
CMU / CMU15-779
Advanced Topics in Machine Learning Systems (LLM Edition)
MIT / MIT6.S978
Deep Generative Models
MIT / MIT6.S184
Generative AI with Stochastic Differential Equations
Stanford / CS486
Advanced Large Language Models Research Seminar
Stanford / CS525
Training Data for AI
PDF 讲义
仅列出已有 PDF；完整讲义请到 Notes 阅读
CME295 / Week 0 · Day 1
第 0 周 · Day 1：张量与自动微分
CME295 / Week 0 · Day 2
第 0 周 · Day 2：优化与训练动力学
CME295 / Week 0 · Day 3
第 0 周 · Day 3：词的表示
CME295 / Week 0 · Day 4
第 0 周 · Day 4：序列模型与注意力的诞生
CME295 / Week 0 · Day 5
第 0 周 · Day 5：字符级语言模型实战 + Week 0 整合