Portfolio ↗

Deep Learning Building Blocks

Convolutions, recurrence, normalisation, activations, optimisers and regularisation.

13concepts
112flashcards
102minutes of reading
Start with Dropout and Modern Regularisation Drill the deck
All levels beginner intermediate advanced
  1. 01 Dropout and Modern Regularisation Why dropout was the dominant regulariser for a decade and why modern LLM training mostly skips it in favour of letting data do the work. beginner 6m 4 cards