Conv2D layers in TensorFlow - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to create a Conv2D layer with 32 filters.

TensorFlow

layer = tf.keras.layers.Conv2D([1], kernel_size=(3,3), activation='relu')

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

A16

B64

C32

D128

Attempts:

3 left

💡 Hint

Common Mistakes

Choosing a very large number of filters can slow training.

Using zero or negative numbers causes errors.

The number of filters in a Conv2D layer defines how many feature detectors it has. Here, 32 filters is the correct choice.

2fill in blank

medium

Complete the code to add padding to the Conv2D layer to keep the output size same as input.

TensorFlow

layer = tf.keras.layers.Conv2D(32, kernel_size=(3,3), padding=[1], activation='relu')

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

A'valid'

B'same'

C'full'

D'none'

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'valid' padding reduces output size.

Using invalid strings causes errors.

Padding='same' adds zeros around the input so the output has the same width and height as the input.

3fill in blank

hard

Fix the error in the Conv2D layer that causes a shape mismatch.

TensorFlow

layer = tf.keras.layers.Conv2D(64, kernel_size=[1], activation='relu')

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

A(3, 3)

B3

C(5, 5)

D[3, 3]

Attempts:

3 left

💡 Hint

Common Mistakes

Passing a single integer instead of a tuple.

Using square brackets instead of parentheses.

The kernel_size must be a tuple of two integers, like (3, 3). Using a single integer or list causes errors.

4fill in blank

hard

Fill both blanks to create a Conv2D layer with 16 filters and stride of 2.

TensorFlow

layer = tf.keras.layers.Conv2D([1], kernel_size=(3,3), strides=[2], activation='relu')

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

A16

B(1, 1)

C(2, 2)

D32

Attempts:

3 left

💡 Hint

Common Mistakes

Using stride (1, 1) does not reduce output size.

Choosing too many filters can slow training.

16 filters and strides of (2, 2) reduce the output size by skipping pixels.

5fill in blank

hard

Fill all three blanks to create a Conv2D layer with 64 filters, kernel size (5,5), and 'valid' padding.

TensorFlow

layer = tf.keras.layers.Conv2D([1], kernel_size=[2], padding=[3], activation='relu')

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

A32

B(5, 5)

C'valid'

D64

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'same' padding when 'valid' is required.

Passing kernel size as a single integer.

64 filters, kernel size (5,5), and 'valid' padding means no padding, so output shrinks.