Code icon

The App is Under a Quick Maintenance

We apologize for the inconvenience. Please come back later

Menu iconMenu iconUnder the Hood of Large Language Models
Under the Hood of Large Language Models

Project 1: Build a Toy Transformer from Scratch in PyTorch

Learning outcomes

  • You built a working decoder-only Transformer from first principles.
  • You understand token→embedding→attention→FFN→logits end-to-end.
  • You can now iterate: add features, measure effects, and refine.

Learning outcomes

  • You built a working decoder-only Transformer from first principles.
  • You understand token→embedding→attention→FFN→logits end-to-end.
  • You can now iterate: add features, measure effects, and refine.

Learning outcomes

  • You built a working decoder-only Transformer from first principles.
  • You understand token→embedding→attention→FFN→logits end-to-end.
  • You can now iterate: add features, measure effects, and refine.

Learning outcomes

  • You built a working decoder-only Transformer from first principles.
  • You understand token→embedding→attention→FFN→logits end-to-end.
  • You can now iterate: add features, measure effects, and refine.