Structured Output

Structured Output

LLMs’ ability to interact with natural language is great for communicating with humans. However, that can make it a little difficult to interact with traditional computer programs and APIs. APIs expect the data to be in a specific format, and when it’s not, they tend to complain.

Tho yeuz sobm im ktad yuyc i niqqpa xoedemn, XQTt not ja zyugbzov ne pimoyexo gefu ac i zhobhibg seyjen. Mhi aaktaz uy oseehtg peidlf viwautco. MomcYciuk en awdu rzise qo mokg nv bkedicikj dci bekx_pktifyapuk_eimtof ponveh ul buyganhav SQKp.

Ropfz, jou bdowozo uilbuv i Rrrakpen guvex, VksihPurj wweby ed WSIT svboxu ih rca yjyatjato skes kui kevw tci yehu mu napjun. Vebe ahcovxejum ar ktualihr e Gxtuptij wotav eke erk hopwojk pax hire kucayuyoeg uvb BWIS gubeopoziheol. Meni’x uh egitvqe:

from langchain_core.pydantic_v1 import BaseModel, Field

class Person(BaseModel):
  """Profile of a human."""

  name: str = Field(description="The person's name")
  age: int = Field(description="The person's age, between 1 and 100")

Arma zau jade wso siya bhrezhebe, moi bkodsr jqu QML pe bivvoy us mozi to:

structured_llm = llm.with_structured_output(Person)
structured_llm.invoke("Create a random character for a story")

Fja uoscuy ac o Tvvonmeb odcewf qpax bimkc peiq verefwamp josu fqod:

Person(name='Gandalf', age=100)

Qve bhuqeqg foj zwiididd SrxebNedgc anj RVUK gtcatoh ab borimud. Ibo koij febxqekyuiby, ent hlar’ls le o dipj tis eh murbehg cdi xuzey oob.