LangChain agents in Prompt Engineering / GenAI - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to create a basic LangChain agent using an OpenAI LLM.

Prompt Engineering / GenAI

from langchain.agents import initialize_agent
from langchain.llms import OpenAI

llm = OpenAI(temperature=0)
agent = initialize_agent(llm, tools=[], agent= [1] )

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

A"zero-shot-react-description"

B"chat-completion"

C"simple-agent"

D"tool-using-agent"

Attempts:

3 left

💡 Hint

Common Mistakes

Using an invalid agent type string.

Confusing agent types that require tools with those that don't.

The 'zero-shot-react-description' agent type allows the agent to reason and act without prior examples, suitable for basic LangChain agents.

2fill in blank

medium

Complete the code to add a tool to the LangChain agent's tool list.

Prompt Engineering / GenAI

from langchain.agents import initialize_agent
from langchain.llms import OpenAI
from langchain.tools import Tool

def dummy_tool_func(input_text):
    return "Processed: " + input_text

tool = Tool(name="DummyTool", func=[1], description="A dummy tool.")

llm = OpenAI(temperature=0)
agent = initialize_agent(llm, tools=[tool], agent="zero-shot-react-description")

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

Alambda x: x

Bdummy_tool_func()

C"dummy_tool_func"

Ddummy_tool_func

Attempts:

3 left

💡 Hint

Common Mistakes

Passing the function call instead of the function object.

Passing the function name as a string.

The 'func' parameter expects a function object, so we pass the function name without parentheses.

3fill in blank

hard

Fix the error in the code to run the LangChain agent with an input query.

Prompt Engineering / GenAI

response = agent.[1]("What is the capital of France?")

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

Arespond

Brun

Ccall

Dexecute

Attempts:

3 left

💡 Hint

Common Mistakes

Using non-existent method names like 'execute' or 'call'.

Confusing agent methods with other API methods.

The correct method to run the agent on an input is 'run'. Other method names do not exist on the agent object.

4fill in blank

hard

Fill both blanks to create a tool that returns the length of the input text and add it to the agent.

Prompt Engineering / GenAI

from langchain.tools import Tool

def length_tool_func(text):
    return len(text)

tool = Tool(name=[1], func=[2], description="Returns length of input text.")

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

A"LengthTool"

Blength_tool_func

C"TextLength"

Dlen

Attempts:

3 left

💡 Hint

Common Mistakes

Passing the built-in 'len' function instead of the custom function.

Using a non-string for the tool name.

The tool name should be a string like 'LengthTool', and the function should be the function object 'length_tool_func'.

5fill in blank

hard

Fill all three blanks to create a LangChain agent with a tool that reverses input text and run it on a query.

Prompt Engineering / GenAI

from langchain.agents import initialize_agent
from langchain.llms import OpenAI
from langchain.tools import Tool

def reverse_text(text):
    return text[::-1]

reverse_tool = Tool(name=[1], func=[2], description="Reverses input text.")

llm = OpenAI(temperature=0)
agent = initialize_agent(llm, tools=[reverse_tool], agent=[3])

result = agent.run("Hello World")

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

A"ReverseTool"

Breverse_text

C"zero-shot-react-description"

D"reverse-agent"

Attempts:

3 left

💡 Hint

Common Mistakes

Using incorrect agent type strings.

Passing function names as strings instead of function objects.

The tool name is 'ReverseTool', the function is the custom 'reverse_text', and the agent type is 'zero-shot-react-description' for basic reasoning.