Deepwatch

Courses, Andrej Karpathy

Neural Networks: Zero to Hero

10 lectures, 19 hr 21 min in all

0 of 10 watched

Start lecture 1
  1. 1
    The spelled-out intro to neural networks and backpropagation: building micrograd

    2 hr 26

  2. 2
    The spelled-out intro to language modeling: building makemore

    1 hr 58

  3. 3
    Building makemore Part 2: MLP

    1 hr 16

  4. 4
    Building makemore Part 3: Activations & Gradients, BatchNorm

    1 hr 56

  5. 5
    Building makemore Part 4: Becoming a Backprop Ninja

    1 hr 55

  6. 6
    Building makemore Part 5: Building a WaveNet

    56 min

  7. 7
    Let's build GPT: from scratch, in code, spelled out.

    1 hr 56

  8. 8
    State of GPT | BRK216HFS

    43 min

  9. 9
    Let's build the GPT Tokenizer

    2 hr 14

  10. 10
    Let's reproduce GPT-2 (124M)

    4 hr 01