Image dataset from folders in PyTorch - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the PyTorch image dataset loader.

PyTorch

from torchvision.datasets import [1]

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

AImageFolder

BDataLoader

Ctransforms

Ddatasets

Attempts:

3 left

💡 Hint

Common Mistakes

Using DataLoader instead of ImageFolder for loading images.

Importing transforms instead of ImageFolder.

The ImageFolder class loads images from folders arranged by class.

2fill in blank

medium

Complete the code to create an ImageFolder dataset from the 'data/train' folder.

PyTorch

dataset = ImageFolder(root='[1]')

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

Atrain_data

Bdata/test

Cdata/train

Dimages/train

Attempts:

3 left

💡 Hint

Common Mistakes

Using the test folder path instead of train.

Using a wrong folder name.

The root parameter should point to the folder containing class subfolders, here 'data/train'.

3fill in blank

hard

Fix the error in the code to apply transforms to the ImageFolder dataset.

PyTorch

from torchvision import transforms

transform = transforms.Compose([transforms.Resize(256), transforms.CenterCrop(224), transforms.ToTensor()])
dataset = ImageFolder(root='data/train', [1]=transform)

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

Atransformer

Btransforms

Ctransformations

Dtransform

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'transforms' instead of 'transform' as parameter name.

Using 'transformations' or 'transformer' which are invalid.

The correct argument name is transform to apply transformations to the dataset.

4fill in blank

hard

Fill both blanks to create a DataLoader for the dataset with batch size 32 and shuffling enabled.

PyTorch

from torch.utils.data import [1]
dataloader = [2](dataset, batch_size=32, shuffle=True)

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

ADataLoader

BDataset

Cdataloader

DLoader

Attempts:

3 left

💡 Hint

Common Mistakes

Importing Dataset instead of DataLoader.

Using lowercase 'dataloader' which is not a class name.

The DataLoader class is imported and used to create batches and shuffle the dataset.

5fill in blank

hard

Fill all three blanks to print the class names from the ImageFolder dataset.

PyTorch

class_names = dataset.[1]
for i, class_name in enumerate(class_names):
    print(f"Class [2]: [3]")

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

Aclasses

Bi

Cclass_name

Dlabels

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'labels' instead of 'classes' attribute.

Printing wrong variables inside the loop.

The classes attribute holds class names. The loop prints index i and name class_name.