Dataset from files in TensorFlow - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to load a dataset from a CSV file using TensorFlow.

TensorFlow

dataset = tf.data.experimental.make_csv_dataset('data.csv', batch_size=[1])

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

A1

B10

C100

D32

Attempts:

3 left

💡 Hint

Common Mistakes

Using batch size 1 can be very slow.

Using too large batch size can cause memory errors.

The batch size controls how many samples are loaded at once. 32 is a common batch size.

2fill in blank

medium

Complete the code to shuffle the dataset with a buffer size of 1000.

TensorFlow

dataset = dataset.shuffle(buffer_size=[1])

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

A1000

B100

C10

D5000

Attempts:

3 left

💡 Hint

Common Mistakes

Using too small buffer size results in poor shuffling.

Using too large buffer size may use too much memory.

Shuffling with a buffer size of 1000 ensures good mixing of data.

3fill in blank

hard

Fix the error in the code to map a function that normalizes features.

TensorFlow

def normalize(features, label):
    features = tf.cast(features, tf.float32) / [1]
    return features, label

dataset = dataset.map(normalize)

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

A255.0

B1.0

C100.0

D0.255

Attempts:

3 left

💡 Hint

Common Mistakes

Dividing by 1.0 does not change values.

Dividing by 0.255 is incorrect scaling.

Dividing by 255.0 normalizes pixel values from 0-255 to 0-1.

4fill in blank

hard

Fill both blanks to batch the dataset and repeat it indefinitely.

TensorFlow

dataset = dataset.[1](batch_size=32).[2]()

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

Abatch

Bshuffle

Crepeat

Dmap

Attempts:

3 left

💡 Hint

Common Mistakes

Calling repeat before batch changes the dataset shape.

Using shuffle instead of batch changes data order but not batching.

First batch the dataset, then repeat it indefinitely for training.

5fill in blank

hard

Fill all three blanks to create a dataset from text files, shuffle, and batch it.

TensorFlow

files = tf.data.Dataset.list_files([1])
dataset = files.interleave(tf.data.TextLineDataset, cycle_length=4)
dataset = dataset.[2](buffer_size=1000).[3](batch_size=64)

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

A'data/*.txt'

Bshuffle

Cbatch

D'data.csv'

Attempts:

3 left

💡 Hint

Common Mistakes

Using a single file name instead of a pattern.

Batching before shuffling reduces randomness.

Use list_files with a pattern to find text files, then shuffle and batch the dataset.