Instructor

import instructor
import weave
from pydantic import BaseModel
from openai import OpenAI


# 원하는 출력 구조 정의
class UserInfo(BaseModel):
    user_name: str
    age: int

# Weave 초기화
weave.init(project_name="instructor-test")

# OpenAI 클라이언트 패치
client = instructor.from_openai(OpenAI())

# 자연어에서 구조화된 데이터 추출
user_info = client.chat.completions.create(
    model="gpt-3.5-turbo",
    response_model=UserInfo,
    messages=[{"role": "user", "content": "John Doe is 30 years old."}],
)

import instructor
import weave
from openai import OpenAI
from pydantic import BaseModel


# 원하는 출력 구조 정의
class Person(BaseModel):
    person_name: str
    age: int


# Weave 초기화
weave.init(project_name="instructor-test")

# OpenAI 클라이언트 패치
lm_client = instructor.from_openai(OpenAI())


# 자연어에서 구조화된 데이터 추출
@weave.op()
def extract_person(text: str) -> Person:
    return lm_client.chat.completions.create(
        model="gpt-3.5-turbo",
        messages=[
            {"role": "user", "content": text},
        ],
        response_model=Person,
    )


person = extract_person("My name is John and I am 20 years old")

import asyncio
from typing import List, Iterable

import instructor
import weave
from openai import AsyncOpenAI
from pydantic import BaseModel


# 원하는 출력 구조 정의
class Person(BaseModel):
    person_name: str
    age: int


# Weave 초기화
weave.init(project_name="instructor-test")

# OpenAI 클라이언트 패치
lm_client = instructor.from_openai(AsyncOpenAI())


class PersonExtractor(weave.Model):
    openai_model: str
    max_retries: int

    @weave.op()
    async def predict(self, text: str) -> List[Person]:
        model = await lm_client.chat.completions.create(
            model=self.openai_model,
            response_model=Iterable[Person],
            max_retries=self.max_retries,
            stream=True,
            messages=[
                {
                    "role": "system",
                    "content": "You are a perfect entity extraction system",
                },
                {
                    "role": "user",
                    "content": f"Extract `{text}`",
                },
            ],
        )
        return [m async for m in model]


model = PersonExtractor(openai_model="gpt-4", max_retries=2)
asyncio.run(model.predict("John is 30 years old"))