Skip to content
Learn with AlakhSharma

C3

Optimization for ML

5 lessons2.5 hoursPlannedFree

Outcome

Implement and diagnose gradient-based optimizers.

Prerequisites

Lessons

  1. L01Gradient descent and the loss landscape · coming soonExercise: Optimize a 2D function and plot the path.
  2. L02Learning rate effects · coming soonExercise: Sweep learning rates and report divergence.
  3. L03Stochastic gradient descent and mini-batches · coming soonExercise: Compare batch sizes on noise and speed.
  4. L04Momentum and Adam · coming soonExercise: Implement Adam from the equations.
  5. L05Diagnosing a run that won't converge · coming soonExercise: A debugging checklist applied to a broken script.