StepLR and MultiStepLR in PyTorch - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to create a StepLR scheduler that decreases the learning rate every 5 epochs.

PyTorch

scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=[1], gamma=0.1)

Drag options to blanks, or click blank then click option'

A10

B5

C3

D7

Attempts:

3 left

💡 Hint

Common Mistakes

Using a wrong step_size value that doesn't match the desired interval.

The step_size parameter controls how often the learning rate decreases. Here, it should be 5 to reduce every 5 epochs.

2fill in blank

medium

Complete the code to create a MultiStepLR scheduler that decreases the learning rate at epochs 10 and 20.

PyTorch

scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer, milestones=[1], gamma=0.1)

Drag options to blanks, or click blank then click option'

A[12, 22]

B[5, 15]

C[8, 18]

D[10, 20]

Attempts:

3 left

💡 Hint

Common Mistakes

Using wrong milestone epochs that don't match the schedule.

The milestones parameter is a list of epochs where the learning rate will be reduced. Here, it should be [10, 20].

3fill in blank

hard

Fix the error in the code to correctly update the learning rate scheduler after each epoch.

PyTorch

for epoch in range(num_epochs):
    train()
    validate()
    [1]

Drag options to blanks, or click blank then click option'

Ascheduler.step()

Bscheduler.update()

Cscheduler.adjust()

Dscheduler.reset()

Attempts:

3 left

💡 Hint

Common Mistakes

Using non-existent methods like update(), adjust(), or reset().

The correct method to update the learning rate scheduler after each epoch is scheduler.step().

4fill in blank

hard

Fill both blanks to create a StepLR scheduler with a step size of 7 and a decay factor of 0.5.

PyTorch

scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=[1], gamma=[2])

Drag options to blanks, or click blank then click option'

A7

B0.1

C0.5

D5

Attempts:

3 left

💡 Hint

Common Mistakes

Mixing up step_size and gamma values.

The step_size should be 7 and the gamma (decay factor) should be 0.5 as specified.

5fill in blank

hard

Fill all three blanks to create a MultiStepLR scheduler with milestones at epochs 8 and 16, and a decay factor of 0.2.

PyTorch

scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer, milestones=[1], gamma=[2])

for epoch in range(num_epochs):
    train()
    validate()
    [3]

Drag options to blanks, or click blank then click option'

A[8, 16]

Bscheduler.step()

C0.2

D0.5

Attempts:

3 left

💡 Hint

Common Mistakes

Using wrong milestone epochs, wrong gamma, or forgetting to call scheduler.step().

Milestones are [8, 16], gamma is 0.2, and the scheduler is updated each epoch with scheduler.step().