Streaming responses in Prompt Engineering / GenAI - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to start streaming the AI model's response.

Prompt Engineering / GenAI

response = model.generate_stream(prompt=[1])

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

Atext

Binput_text

Cprompt

Dquery

Attempts:

3 left

💡 Hint

Common Mistakes

Using variable names like 'text' or 'query' which are not defined in this context.

The method generate_stream requires the prompt parameter to specify the input text for streaming the response.

2fill in blank

medium

Complete the code to read the streamed tokens from the response.

Prompt Engineering / GenAI

for token in response.[1]():
    print(token)

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

Astream_tokens

Bstream

Citer_tokens

Dread

Attempts:

3 left

💡 Hint

Common Mistakes

Using generic method names like 'read' or 'stream' which do not exist.

The method stream_tokens() is used to iterate over tokens as they are streamed from the model.

3fill in blank

hard

Fix the error in the code to properly handle streaming with a callback function.

Prompt Engineering / GenAI

def on_token(token):
    print(token)

model.generate_stream(prompt=prompt, [1]=on_token)

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

Acallback

Bon_token

Cstream_callback

Dtoken_handler

Attempts:

3 left

💡 Hint

Common Mistakes

Passing the function with incorrect parameter names like 'on_token' or 'token_handler'.

The parameter to pass a function to handle tokens during streaming is commonly named 'callback'.

4fill in blank

hard

Fill both blanks to correctly initialize streaming with a timeout and a callback.

Prompt Engineering / GenAI

response = model.generate_stream(prompt=prompt, [1]=5, [2]=handle_token)

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

Atimeout

Bmax_tokens

Ccallback

Dstream

Attempts:

3 left

💡 Hint

Common Mistakes

Confusing 'max_tokens' with 'timeout' or using wrong parameter names.

The 'timeout' parameter sets the max wait time, and 'callback' is the function to handle tokens during streaming.

5fill in blank

hard

Fill both blanks to collect streamed tokens into a string with a callback and print final output.

Prompt Engineering / GenAI

collected = []
def collect_token(token):
    collected.[1](token)

model.generate_stream(prompt=prompt, [2]=collect_token)
print(''.join(collected))

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

Aappend

Bcallback

D.strip()

Attempts:

3 left

💡 Hint

Common Mistakes

Using wrong list method like 'extend' or adding extra string methods after join.

Use 'append' to add tokens to the list, 'callback' to pass the function, and no extra method after join to print raw output.