Source code for langchain.chains.openai_functions.citation_fuzzy_match
from typing import Iterator, List
from langchain_core.language_models import BaseLanguageModel
from langchain_core.messages import HumanMessage, SystemMessage
from langchain_core.output_parsers.openai_functions import PydanticOutputFunctionsParser
from langchain_core.prompts.chat import ChatPromptTemplate, HumanMessagePromptTemplate
from langchain_core.pydantic_v1 import BaseModel, Field
from langchain.chains.llm import LLMChain
from langchain.chains.openai_functions.utils import get_llm_kwargs
[docs]class FactWithEvidence(BaseModel):
"""代表单个语句的类。
每个事实都有一个主体和一个来源列表。
如果有多个事实,请确保将它们分开,
使得每个事实只使用与其相关的一组来源。"""
fact: str = Field(..., description="Body of the sentence, as part of a response")
substring_quote: List[str] = Field(
...,
description=(
"Each source should be a direct quote from the context, "
"as a substring of the original content"
),
)
def _get_span(self, quote: str, context: str, errs: int = 100) -> Iterator[str]:
import regex
minor = quote
major = context
errs_ = 0
s = regex.search(f"({minor}){{e<={errs_}}}", major)
while s is None and errs_ <= errs:
errs_ += 1
s = regex.search(f"({minor}){{e<={errs_}}}", major)
if s is not None:
yield from s.spans()
[docs] def get_spans(self, context: str) -> Iterator[str]:
for quote in self.substring_quote:
yield from self._get_span(quote, context)
[docs]class QuestionAnswer(BaseModel):
"""一个问题及其答案,作为每个事实的列表,每个事实应该有一个来源。
每个句子包含一个主体和一个来源列表。"""
question: str = Field(..., description="Question that was asked")
answer: List[FactWithEvidence] = Field(
...,
description=(
"Body of the answer, each fact should be "
"its separate object with a body and a list of sources"
),
)
[docs]def create_citation_fuzzy_match_chain(llm: BaseLanguageModel) -> LLMChain:
"""创建一个引文模糊匹配链。
参数:
llm: 用于该链的语言模型。
返回:
可用于回答带有引文的问题的链 (LLMChain)。
"""
output_parser = PydanticOutputFunctionsParser(pydantic_schema=QuestionAnswer)
schema = QuestionAnswer.schema()
function = {
"name": schema["title"],
"description": schema["description"],
"parameters": schema,
}
llm_kwargs = get_llm_kwargs(function)
messages = [
SystemMessage(
content=(
"You are a world class algorithm to answer "
"questions with correct and exact citations."
)
),
HumanMessage(content="Answer question using the following context"),
HumanMessagePromptTemplate.from_template("{context}"),
HumanMessagePromptTemplate.from_template("Question: {question}"),
HumanMessage(
content=(
"Tips: Make sure to cite your sources, "
"and use the exact words from the context."
)
),
]
prompt = ChatPromptTemplate(messages=messages) # type: ignore[arg-type, call-arg]
chain = LLMChain(
llm=llm,
prompt=prompt,
llm_kwargs=llm_kwargs,
output_parser=output_parser,
)
return chain