Functional API basics in TensorFlow - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is the Functional API in TensorFlow?

The Functional API is a way to build neural networks by defining layers as functions and connecting them. It allows creating complex models with multiple inputs and outputs, unlike the simple sequential model.

beginner

How do you define an input layer using the Functional API?

You use tf.keras.Input(shape=(input_shape,)) to create an input layer that specifies the shape of the input data.

beginner

How do you connect layers in the Functional API?

You call a layer like a function on the output of the previous layer. For example, x = tf.keras.layers.Dense(10)(input_layer) connects a Dense layer to the input.

beginner

How do you create a model using the Functional API?

You create a model by specifying inputs and outputs: model = tf.keras.Model(inputs=input_layer, outputs=output_layer).

intermediate

Why use the Functional API instead of Sequential API?

The Functional API supports models with multiple inputs or outputs, shared layers, and non-linear topology, which the Sequential API cannot handle.

Which TensorFlow function is used to define the input layer in the Functional API?

A tf.keras.Input()

B tf.keras.Layer()

C tf.keras.Sequential()

D tf.keras.Dense()

tf.keras.Input() defines the input layer specifying the shape of the input data.

How do you connect layers in the Functional API?

A By stacking layers sequentially only

B By adding layers to a list

C By calling one layer as a function on the output of another layer

D By using a for loop

In the Functional API, layers are connected by calling them as functions on previous layers' outputs.

What is the correct way to create a model in the Functional API?

A tf.keras.Sequential()

B tf.keras.Model(inputs=input_layer, outputs=output_layer)

C tf.keras.Model() without inputs or outputs

D tf.keras.Layer()

You create a model by specifying the input and output layers explicitly.

Which of the following is a key advantage of the Functional API over the Sequential API?

A Supports multiple inputs and outputs

B Easier to use for simple models

C Automatically compiles the model

D Requires less code for sequential models

The Functional API supports complex architectures with multiple inputs and outputs.

What does the following code do? x = Dense(10)(input_layer)

A Creates a new input layer

B Defines the output layer only

C Compiles the model

D Connects a Dense layer with 10 units to the input_layer

This code applies a Dense layer with 10 units to the input_layer, connecting them.

Explain how to build a simple neural network model using the Functional API in TensorFlow.

Think about how layers connect and how the model is finalized.

Describe the benefits of using the Functional API compared to the Sequential API.

Consider model complexity and flexibility.