TensorFlow

Question 1

Which of the following statements is true of TensorFlow?

Answer 1

TensorFlow is a scalable and multi platform programming interface for implementing and
running machine learning algorithms, including convenience wrappers for deep learning.

Question 2

How does TensorFlow represent numeric computations?

Answer 2

Using a Directed Acyclic Graph (or DAG)

Question 3

Which are useful components when building custom Neural Network models?

Answer 3

tf.losses

tf.metrics

tf.optimizers

Question 4

Which API is used to build performant, complex input pipelines from simple, re-usable pieces that will feed your model’s training or evaluation loops.

Answer 4

tf.data.Dataset

Question 5

What operations can be performed on tensors

Answer 5

They can be both reshaped and sliced

Question 6

What is true when we compute a loss gradient? (3 Things)

Answer 6

1. TensorFlow records all operations executed inside the context of a tf.GradientTape onto a
   tape.
2. It uses tape and the gradients associated with each recorded operation to compute the
   gradients.
3. The computed gradient of a recorded computation will be used in reverse mode
   differentiation.

Question 7

What are distinct ways to create a dataset?

Answer 7

Feature columns describe how the model should use raw input data from your features
dictionary.

Question 8

Which of the following is true about embedding?

Answer 8

1. An embedding is a weighted sum of the feature crossed values.
2. Embedding is a handy adapter that allows a network to incorporate spores or categorical data.
3. The number of embeddings is the hyperparameter to your machine learning model.

Question 9

What is the use of tf.keras.layers.TextVectorization?

Answer 9

It turns raw strings into an encoded representation that can be read by an Embedding layer or Dense layer.

Question 10

Which is not a part of Categorical features preprocessing

Answer 10

tf.keras.layers.Discretization

Question 11

Which of the layers is non-trainable?

Answer 11

Hashing

Question 12

When should you avoid using the Keras function adapt()?

Answer 12

When working with lookup layers with very large vocabularies

Question 13

Which is a part of Keras preprocessing layers?

Answer 13

1. Image data augmentation
2. Image preprocessing
3. Numerical features preprocessing

Question 14

True or false: Non-linearity helps in training your model at a much faster rate and with more accuracy without the loss of your important information?

Answer 14

True

Question 15

During the training process, each additional layer in your network can successively reduce signal vs. noise. How can we fix this?

Answer 15

Use non-saturating, nonlinear activation functions such as ReLUs.

Question 16

How does Adam (optimization algorithm) help in compiling the Keras model?

Answer 16

Both by updating network weights iteratively based on training data by diagonal rescaling of
the gradients

Question 17

The predict function in the tf.keras API returns what?

Answer 17

Numpy array(s) of predictions

Question 18

What is the significance of the Fit method while training a Keras model?

Answer 18

Defines the number of epochs

Question 19

Difference to Keras Functional API

Answer 19

Unlike the Keras Sequential API, we have to provide the shape of the input to the model.

Question 20

The Keras Functional API can be characterized by having

Answer 20

Multiple inputs and outputs and models with shared layers

Question 21

How does regularization help build generalizable models ?

Answer 21

By adding dropout layers to our neural networks

Question 22

The L2 regularization provides what?

Answer 22

It adds a sum of the squared parameter weights term to the loss function.

Question 23

. When sending training jobs to Vertex AI, it is common to split most of the logic into a _________ and a ___________ file

Answer 23

task.py, model.py

Question 24

Which file is the entry point to your code that Vertex AI will start and contains details such as“how to parse command-line arguments and where to write model outputs?

Answer 24

task.py

Question 25

When you package up a TensorFlow model as a Python Package, what statement should every
Python module contain in every folder?

Answer 25

an init_.py

Question 26

To make your code compatible with Verte AI, there are three basic steps that must be completed in a specific order.

Answer 26

First, upload data to Google Cloud Storage. Next, move code into a trainer Python
package. Then submit your training job with gcloud to train on Vertex AI.

Question 27

You can use either pre-built containers or custom containers to run training
jobs. Both containers require you specify settings that Vertex AI needs to run your training code,
including __________, ____________, and ________.

Answer 27

Region, display-name, worker-pool-spec