Fill-in the blanks in the following statement with one of the propositions below:…
- User
Fill-in the blanks in the following statement with one of the propositions below: Interacting with Large Language Models (LLMs) differs from traditional machine learning models. Working with LLMs involves natural language input, known as a _____, resulting in output from the Large Language Model, known as the ______ .
- Assistant
Interacting with Large Language Models (LLMs) differs from traditional machine learning models. Working with LLMs involves natural language input, known as a prompt, resulting in output from the Large Language Model, known as the completion.
- User
True or False – Is the following statement True or False? "RNNs are better at generating text than Transformers."
- Assistant
False.
Transformers have largely replaced RNNs for text generation tasks because they handle long-range dependencies better, are more parallelizable, and generally produce higher-quality text. Models like GPT, which are based on the Transformer architecture, have set new standards in text generation.
- User
Which of the following propositions best describes the self-attention mechanism that powers the transformer architecture? Only 1 correct answer to select
The ability of the transformer to analyze its own performance and make adjustments accordingly.
A measure of how well a model can understand and generate human-like language.
A mechanism that allows a model to focus on different parts of the input, enabling it to learn the relevance and context of all of the words in a sentence.
A technique used to improve the generalization capabilities of a model by training it on diverse datasets.
- Assistant
The correct answer is:
A mechanism that allows a model to focus on different parts of the input, enabling it to learn the relevance and context of all of the words in a sentence.
- User
Among the following propositions, which one best describes tokenization in the context of the Transformer architecture? Only 1 correct answer to select
Tokenization is the process of converting a sequence of words into a single token for efficient computation.
Tokenization is the process of splitting a text into words (or pieces of words) and then converting them into numbers.
Tokenization is the process of translating text from one language to another.
Tokenization is the process of applying sentiment analysis to text data.
- Assistant
The correct answer is:
Tokenization is the process of splitting a text into words (or pieces of words) and then converting them into numbers.
- User
Which of the following transformer-based models is well-suited to the task of text translation? Only 1 correct answer to select
Autoregressive models
Encoder-decoder models
Autoencoder models
Autonomous models
- Assistant
The correct answer is:
Encoder-decoder models
- User
Which of the following propositions refers to the area where the users inputs a prompt to interact with a Large Language Model? Only 1 correct answer to select
The context area
The prompt
The chat
The context window
- Assistant
The correct answer is:
The prompt
- User
Which of the following propositions refers to the in-context learning method for which the user includes in the prompt a single example of the task to perform (question and answer) as well as the actual task to be performed by the LLM? Only 1 correct answer to select
Few shots
Hot shot
One shot
Zero shot
- Assistant
The correct answer is:
One shot
- User
Among the following propositions, which one is the LLM configuration parameter that can be adjusted to either increase or decrease randomness of the model output? Only 1 correct answer to select
Max new tokens
Top-k sampling
Temperature
Number of beams & beam search
- Assistant
The correct answer is:
Temperature
- User
Which of the following stages are part of the generative AI model lifecycle discussed in the course? 4 correct answers to select
Selecting a candidate model (or potentially pre-training a custom model)
Defining the problem scope and use-cases
Performing regularization
Manipulating the model to improve performance
Deploying the model and integrating it in applications
- Assistant
The four correct answers are:
- Selecting a candidate model (or potentially pre-training a custom model)
- Defining the problem scope and use-cases
- Manipulating the model to improve performance
- Deploying the model and integrating it in applications
- User
Which of the following propositions is a critical step in the GenAI Project Lifecycle? Only 1 correct answer to select
Model fine-tuning
Model evaluation
Model customization
Front-end design
- Assistant
The correct answer is:
Model evaluation
- User
What are some additional courses you would like to see from DeepLearning.AI?
- Assistant
As an AI assistant, I don’t have personal preferences, but based on community interest and trends in AI and machine learning, here are some course ideas that many learners and professionals might find valuable from DeepLearning.AI:
-
Ethics and Responsible AI Development
Covering bias mitigation, fairness, transparency, and ethical considerations in AI. -
Advanced Reinforcement Learning
Diving deeper into modern RL techniques, multi-agent RL, and real-world applications. -
Deploying and Scaling AI Models
Focused on practical deployment, model serving, optimization, and scaling in production environments. -
AI for Healthcare
Applying machine learning and deep learning in medical imaging, diagnostics, and personalized medicine. -
Computer Vision Beyond Basics
Covering topics like video understanding, 3D vision, and advanced techniques like attention in vision models. -
Natural Language Understanding (NLU) & Dialogue Systems
Building conversational agents, chatbots, and intent recognition beyond just language models. -
Machine Learning Engineering
Best practices for building reliable, maintainable ML systems, MLOps, and continuous integration.
If you have specific interests, I can also help tailor recommendations!
-
- User
Question 1 Interacting with Large Language Models (LLMs) differs from traditional machine learning models. Working with LLMs involves natural language input, known as a _____, resulting in output from the Large Language Model, known as the ______ .
Choose the answer that correctly fill in the blanks.
tunable request, completion
prompt, fine-tuned LLM
prediction request, prediction response
prompt, completion
1 point
- Assistant
The correct answer is:
prompt, completion
- User
Question 2 Large Language Models (LLMs) are capable of performing multiple tasks supporting a variety of use cases. Which of the following tasks supports the use case of converting code comments into executable code?
Translation
Invoke actions from text
Text summarization
Information Retrieval
1 point
- Assistant
The correct answer is:
Invoke actions from text
- User
Question 3 What is the self-attention that powers the transformer architecture?
The ability of the transformer to analyze its own performance and make adjustments accordingly.
A measure of how well a model can understand and generate human-like language.
A technique used to improve the generalization capabilities of a model by training it on diverse datasets.
A mechanism that allows a model to focus on different parts of the input sequence during computation.
1 point
- Assistant
The correct answer is:
A mechanism that allows a model to focus on different parts of the input sequence during computation.
- User
Which of the following stages are part of the generative AI model lifecycle mentioned in the course? (Select all that apply)
Selecting a candidate model and potentially pre-training a custom model.
Defining the problem and identifying relevant datasets.
Deploying the model into the infrastructure and integrating it with the application.
Manipulating the model to align with specific project needs.
Performing regularization
1 point
- Assistant
The correct answers are:
- Selecting a candidate model and potentially pre-training a custom model.
- Defining the problem and identifying relevant datasets.
- Deploying the model into the infrastructure and integrating it with the application.
- Manipulating the model to align with specific project needs.
- User
Question 5 "RNNs are better than Transformers for generative AI Tasks."
Is this true or false?
True
False
1 point
- Assistant
The correct answer is:
False
- User
Question 6 Which transformer-based model architecture has the objective of guessing a masked token based on the previous sequence of tokens by building bidirectional representations of the input sequence.
Autoregressive
Autoencoder
Sequence-to-sequence
1 point
- Assistant
The correct answer is:
Autoencoder
- User
Question 7 Which transformer-based model architecture is well-suited to the task of text translation?
Autoencoder
Sequence-to-sequence
Autoregressive
1 point
- Assistant
The correct answer is:
Sequence-to-sequence
- User
Question 8 Do we always need to increase the model size to improve its performance?
True
False
1 point
- Assistant
The correct answer is:
False
- User
Scaling laws for pre-training large language models consider several aspects to maximize performance of a model within a set of constraints and available scaling choices. Select all alternatives that should be considered for scaling when performing model pre-training?
Model size: Number of parameters
Compute budget: Compute constraints
Dataset size: Number of tokens
Batch size: Number of samples per iteration
1 point
- Assistant
The correct answers are:
- Model size: Number of parameters
- Compute budget: Compute constraints
- Dataset size: Number of tokens
- Batch size: Number of samples per iteration
- User
Question 10 "You can combine data parallelism with model parallelism to train LLMs."
Is this true or false?
True
False
1 point
- Assistant
The correct answer is:
True