Learning path
AI from Scratch
Implement the math behind modern ML — backprop in a neural network, then a GPT-style transformer. No PyTorch shortcuts.
advanced2 courses39 lessons~10 hoursFree
Start pathor sign up to track progress
Course sequence
- 01
Build a Neural Network
Not startedBuild an MNIST classifier from scratch in NumPy: neurons, dense layers, sigmoid/ReLU/softmax, MSE/cross-entropy loss, manual backpropagation, gradient descent + Adam, training loop, and the path to PyTorch/TensorFlow.
advanced25 lessons~6h - 02
Build a Transformer (GPT-style)
Not startedBuild a tiny GPT in NumPy: token + position embeddings, layer normalization, scaled dot-product self-attention, multi-head attention with causal mask, feed-forward networks, residual connections, sampling (greedy, top-k, top-p), and the architecture behind ChatGPT, Claude, Llama.
advanced14 lessons~4h
Create a free account to track your progress across this path.