OpenAI embeddings API in Prompt Engineering / GenAI - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

What is the output of this OpenAI embeddings API call?

Consider this Python code snippet that calls the OpenAI embeddings API to get vector embeddings for the text 'Hello world'. What will be the type of the variable embedding after running this code?

Prompt Engineering / GenAI

from openai import OpenAI
client = OpenAI()
response = client.embeddings.create(
    model="text-embedding-3-large",
    input="Hello world"
)
embedding = response.data[0].embedding
print(type(embedding))

A <class 'float'>

B <class 'dict'>

C <class 'str'>

D <class 'list'>

💡 Hint

The embedding is a vector of numbers representing the input text.

The OpenAI embeddings API returns a list of floats representing the embedding vector. So the type of embedding is a list.

🧠 Conceptual

intermediate

Which option best describes the purpose of embeddings in AI?

Embeddings are used in AI to convert text or other data into a form that machines can understand. Which of the following best describes what embeddings do?

A They convert input data into fixed-length vectors capturing semantic meaning.

B They generate human-readable summaries of text.

C They translate text from one language to another.

D They classify input data into predefined categories.

💡 Hint

Think about how machines understand text as numbers.

Embeddings transform data into fixed-length numeric vectors that capture meaning, allowing machines to compare and analyze data effectively.

❓ Hyperparameter

advanced

Which parameter controls the model choice in the OpenAI embeddings API call?

In the OpenAI embeddings API, you specify which model to use for generating embeddings. Which parameter in the API call sets this model?

A max_tokens

B model

C input

D temperature

💡 Hint

This parameter names the specific embedding model to use.

The 'model' parameter specifies which embedding model the API should use to generate embeddings.

❓ Metrics

advanced

How can you measure similarity between two embeddings?

After obtaining two embedding vectors from the OpenAI embeddings API, which method is commonly used to measure how similar the two vectors are?

A Cosine similarity

B Mean squared error

C Cross entropy loss

D Euclidean distance squared

💡 Hint

Similarity measures how close two vectors point in the same direction.

Cosine similarity measures the cosine of the angle between two vectors, indicating how similar their directions are, which is ideal for embeddings.

🔧 Debug

expert

What error will this OpenAI embeddings API call raise?

Consider this code snippet calling the OpenAI embeddings API with an empty input list. What error will it raise?

Prompt Engineering / GenAI

from openai import OpenAI
client = OpenAI()
response = client.embeddings.create(
    model="text-embedding-3-large",
    input=[]
)

A KeyError

B TypeError

C openai.error.BadRequestError

D No error, returns empty embedding list

💡 Hint

The API expects non-empty input text to generate embeddings.

The OpenAI embeddings API raises a BadRequestError when the input is empty because it requires text to process.