Exponential LR Schedule

Problem

Compute the learning rate at a given step using exponential decay:

lr(step) = initial_lr * decay_rate^step

case 1

in: [0.1,0.95,10]

expected: 0.05987369392368379

case 2

in: [0.01,0.9,0]

expected: 0.01

case 3

in: [1,0.5,5]

expected: 0.03125

case 4

in: [0.1,1,100]

expected: 0.1

Starter codePython

def exponential_lr_schedule(initial_lr, decay_rate, step):
    # Your code here
    pass

4 test cases · run your code against them in the editor