PKU Class 2025 Spring: Introduction to Foundation Models

Instructor: Kun Yuan (kunyuan@pku.edu.cn)

Teaching assistants:

• Jie Hu (hujie@stu.pku.edu.cn)
  
• Yipeng Hu (2301213082@stu.pku.edu.cn)
  
• Qiulin Shang (2100013145@stu.pku.edu.cn)
  
• Yilong Song (2301213059@pku.edu.cn)
  

Office hour: 4pm - 5pm Wednesday, 静园六院220

References

Stanford CS224n: Natural Language Processing with Deep Learning

Lectrures

Lecture 1: Introduction to LLM

• Introduction to large language model [Slides1] [Slides2]
• Reading:
  
  • Andrej Karpathy, State of GPT
    
  • Andrej Karpathy, The busy person’s intro to LLM
    

Lecture 2: Basics in machine learning

• Warm up: Preliminary [Notes]
  
• Linear regression; Logistic regression; Multi-classification; Neural network [Slides]
• Reading:
  
  • Stanford CS231n, Linear classification
    
  • Stanford CS231n, Neural netowrk part I

Lecture 3: Gradient descent

• Convex set; Convex functions; Convex problems; Gradient descent [Slides] [Notes]
• Forward-backward propagation [Notes]

Lecture 4: Stochastic gradient descent

• Stochastic gradient descent (SGD); mini-batch SGD [Slides] [Notes]
  
• Mini-batch forward-backward propagation [Slides]
• Reading:
  
  • Leon Bottou, Large-Scale Machine Learning with Stochastic Gradient Descent
    
  • Yingjie Tian et. al., Recent Advances in Stochastic Gradient Descent in Deep Learning
    

Lecture 5: Adavanced optimizers

• Momentum SGD; Nesterov SGD [Slides]
  
• Adaptive SGD; AdaGrad; RMSProp; Adam [Slides]
  

Lecture 6: Language models

• Word embedding; Language models; Recurrent neural networks [Slides]
  
• Reading:
  
  • Stanford CS224N: Week 3
    

Lecture 7: Transformers

• Seq2seq models; cross-attention; self-attention; transformers [Slides]
  
• Reading:
  
  • Stanford CS224N: Week 4
    
  • Ilya Sutskever et. al., Sequence to Sequence Learning with Neural Networks
    
  • Ashish Vaswani et. al., Attention Is All You Need
    

Lecture 8: Flash Attention

• Memory access cost; Kernal fusion; Flash attention [Slides]

Lecture 9: Midterm Review

• Good Luck [Slides]

Lecture 10: Mixed-Precision

• FP32; FP16; Mixed-precision training [Slides]

Lecture 11: 3D Parallelism

• Data Parallelism; Pipeline Parallelism; Tensor Parallelism [Slides]

Lecture 12: Popular LLM Models

• Teacher forcing; Pretrain and Finetuning; BERT; GPTs [Slides]
• DeepSeek [Slides]
• Reading:
  
  • J. Devlin et.al., BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
  • A. Radford et.al., Improving Language Understanding by Generative Pre-Training
  • A. Radford et.al., Language models are unsupervised multitask learners
  • T. B. Brown et.al., Language Models are Few-Shot Learners
  • DeepSeek, DeepSeek-V3 Technical Report

Lecture 13: Scaling Law

• Scaling law [Slides]
• Reading:
  
  • Jared Kaplan, et. al. Scaling Laws for Neural Language Models
  • Jordan Hoffmann, et. al. Training Compute-Optimal Large Language Models