tf.data.Dataset creation in TensorFlow - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to create a TensorFlow dataset from a list of numbers.

TensorFlow

import tensorflow as tf
numbers = [1, 2, 3, 4, 5]
dataset = tf.data.Dataset.[1](numbers)

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

Afrom_tensor_slices

Bfrom_list

Cfrom_array

Dfrom_data

Attempts:

3 left

💡 Hint

Common Mistakes

Using a non-existent method like from_list or from_array.

Trying to create a dataset directly from a list without using the correct method.

The method from_tensor_slices creates a dataset by slicing the input tensor or list element-wise.

2fill in blank

medium

Complete the code to create a dataset from a range of numbers from 0 to 9.

TensorFlow

import tensorflow as tf
dataset = tf.data.Dataset.[1](10)

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

Arange

Bfrom_list

Cfrom_generator

Dfrom_tensor_slices

Attempts:

3 left

💡 Hint

Common Mistakes

Using from_tensor_slices with an integer instead of a list or tensor.

Using from_generator without defining a generator function.

The range method creates a dataset of numbers from 0 up to (but not including) the given limit.

3fill in blank

hard

Fix the error in the code to create a dataset from a generator function.

TensorFlow

import tensorflow as tf

def gen():
    for i in range(3):
        yield i

dataset = tf.data.Dataset.[1](gen, output_signature=tf.TensorSpec(shape=(), dtype=tf.int32))

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

Arange

Bfrom_generator

Cfrom_tensor_slices

Dfrom_list

Attempts:

3 left

💡 Hint

Common Mistakes

Using from_tensor_slices which expects a tensor or list, not a generator.

Not providing the output_signature argument.

The from_generator method creates a dataset from a Python generator function. It requires an output signature to specify the shape and type of the elements.

4fill in blank

hard

Fill both blanks to create a dataset from a list of tuples and specify the output types.

TensorFlow

import tensorflow as tf

data = [(1, 2.0), (3, 4.0), (5, 6.0)]
dataset = tf.data.Dataset.[1](data, output_signature=(tf.TensorSpec(shape=(), dtype=tf.int32), tf.TensorSpec(shape=(), dtype=[2])))

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

Afrom_generator

Btf.float32

Cfrom_tensor_slices

Dtf.int32

Attempts:

3 left

💡 Hint

Common Mistakes

Using from_generator without a generator function.

Mismatching the output_signature types with the data.

from_tensor_slices creates a dataset by slicing the list of tuples. The output signature must match the types of the tuple elements: first is int32, second is float32.

5fill in blank

hard

Fill all three blanks to create a dataset from a generator with correct output signature for strings.

TensorFlow

import tensorflow as tf

def gen():
    yield "apple"
    yield "banana"
    yield "cherry"

dataset = tf.data.Dataset.[1](gen, output_signature=tf.TensorSpec(shape=(), dtype=[2], name=[3]))

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

Afrom_tensor_slices

Btf.string

C"fruit"

Dfrom_generator

Attempts:

3 left

💡 Hint

Common Mistakes

Using from_tensor_slices instead of from_generator.

Setting dtype to a numeric type instead of tf.string.

Not providing a name or using an invalid name.

from_generator creates a dataset from the generator function. The output signature specifies the shape as scalar (), dtype as tf.string, and a name for the tensor.