Neural Networks: Zero to Hero

Intermediate Computer Science by Andrej Karpathy. 10 lectures, 19.4 hours. Covers: neural networks backpropagation gpt transformers language models pytorch.

Lectures

1. The spelled-out intro to neural networks and backpropagation: building micrograd · 2 h 26 min
2. The spelled-out intro to language modeling: building makemore · 1 h 58 min
3. Building makemore Part 2: MLP · 1 h 16 min
4. Building makemore Part 3: Activations & Gradients, BatchNorm · 1 h 56 min
5. Building makemore Part 4: Becoming a Backprop Ninja · 1 h 55 min
6. Building makemore Part 5: Building a WaveNet · 56 min
7. Let's build GPT: from scratch, in code, spelled out. · 1 h 56 min
8. State of GPT | BRK216HFS · 43 min
9. Let's build the GPT Tokenizer · 2 h 14 min
10. Let's reproduce GPT-2 (124M) · 4 h 1 min

Quick explainers

• But what is a neural network? | Deep learning chapter 1 · 19 min
• A Neural Network Primer · 19 min
• I Built a Neural Network from Scratch · 9 min
• Introducing convolutional neural networks (ML Zero to Hero - Part 3) · 6 min
• Advice for machine learning beginners | Andrej Karpathy and Lex Fridman · 6 min
• Graph Neural Networks Explained: A Clear Guide to GNN Basics & Models · 17 min
• What is a Neural Network? · 8 min
• Neural Networks Explained from Scratch using Python · 18 min
• Natural Language Processing - Tokenization (NLP Zero to Hero - Part 1) · 5 min
• Intro to Machine Learning (ML Zero to Hero - Part 1) · 7 min

All courses