Ai Tip From Hwchase17

Provides AI learning tips by demonstrating how to create custom LangChain agents with structured outputs using Pydantic models.

Description

Automatically generated AI learning skill from curated web and social media sources.

Steps

1. New in LangChain: You can now easily create custom agents with structured outputs!
2. from langchain_core.pydantic_v1 import BaseModel, Field
3. class Joke(BaseModel):
4. setup: str = Field(description="question to set up a joke")
5. punchline: str = Field(description="answer to resolve the joke")
6. llm.with_structured_output(Joke)

Code Examples

from langchain_core.pydantic_v1 import BaseModel, Field

class Joke(BaseModel):
    setup: str = Field(description="question to set up a joke")
    punchline: str = Field(description="answer to resolve the joke")

llm.with_structured_output(Joke)

Dependencies

• Python 3.8+
• Relevant libraries (see code examples)