Compose transforms in PyTorch - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to create a composed transform that converts images to tensors.

PyTorch

transform = transforms.Compose([transforms.To[1]()])

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

ATensor

BImage

CPIL

DArray

Attempts:

3 left

💡 Hint

Common Mistakes

Using ToImage instead of ToTensor.

Using ToPIL which converts tensors back to images.

The ToTensor transform converts a PIL image or numpy array to a PyTorch tensor.

2fill in blank

medium

Complete the code to normalize images with mean 0.5 and std 0.5 using Compose.

PyTorch

transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize(mean=[[1]], std=[0.5])])

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

A0.0

B1.0

C0.5

D0.25

Attempts:

3 left

💡 Hint

Common Mistakes

Using mean 0.0 which does not center the data properly.

Using std 1.0 which does not scale the data.

The mean and std values are set to 0.5 to normalize image pixel values between -1 and 1.

3fill in blank

hard

Fix the error in the Compose transform that should resize images to 128x128 pixels.

PyTorch

transform = transforms.Compose([transforms.Resize([1])])

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

A128

B(128, 128)

C[128, 128]

D128, 128

Attempts:

3 left

💡 Hint

Common Mistakes

Passing a single integer which resizes the smaller edge only.

Using a list instead of a tuple.

The Resize transform expects a tuple for width and height, so (128, 128) is correct.

4fill in blank

hard

Fill both blanks to create a Compose transform that converts images to tensors and then randomly flips them horizontally.

PyTorch

transform = transforms.Compose([transforms.[1](), transforms.[2](p=0.5)])

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

AToTensor

BRandomCrop

CRandomHorizontalFlip

DNormalize

Attempts:

3 left

💡 Hint

Common Mistakes

Using RandomCrop instead of RandomHorizontalFlip.

Normalizing before converting to tensor.

First convert images to tensors with ToTensor, then apply random horizontal flip with RandomHorizontalFlip.

5fill in blank

hard

Fill all three blanks to create a Compose transform that resizes images to 64x64, converts to tensor, and normalizes with mean 0.5 and std 0.5.

PyTorch

transform = transforms.Compose([transforms.Resize([1]), transforms.[2](), transforms.Normalize(mean=[[3]], std=[0.5])])

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

A(64, 64)

BToTensor

C0.5

DRandomVerticalFlip

Attempts:

3 left

💡 Hint

Common Mistakes

Using RandomVerticalFlip instead of ToTensor.

Passing size as a list instead of a tuple.

Resize images to (64, 64), convert to tensor with ToTensor, then normalize with mean and std 0.5.