Optimizer race

Race plain gradient descent, momentum, RMSProp, and Adam from the same start on four loss surfaces.

BeginnerExplained in Better optimizers

Optimizer race

Four optimizers start from the same point on the same surface. Tap the map to move the start.

The Rosenbrock function: easy to find the valley, slow to follow its gently curving floor to the minimum.

Tap to move the start
  • Gradient descent
  • Momentum
  • RMSProp
  • Adam
Surface
1.00x
Racers
Steps0
Gradient descent7.016
Momentum7.016
RMSProp7.016
Adam7.016

Try this

  • Race on the Banana valley with Fast forward, then scale the step sizes to 4x.
  • On the Saddle point, count how many steps each optimizer needs to slide off the ridge.

Try "embedding", "softmax", "overfitting", or "backpropagation".