"""加载问题回答链。"""
from typing import Any, Mapping, Optional, Protocol
from langchain_core.callbacks import BaseCallbackManager, Callbacks
from langchain_core.language_models import BaseLanguageModel
from langchain_core.prompts import BasePromptTemplate
from langchain.chains import ReduceDocumentsChain
from langchain.chains.combine_documents.base import BaseCombineDocumentsChain
from langchain.chains.combine_documents.map_reduce import MapReduceDocumentsChain
from langchain.chains.combine_documents.map_rerank import MapRerankDocumentsChain
from langchain.chains.combine_documents.refine import RefineDocumentsChain
from langchain.chains.combine_documents.stuff import StuffDocumentsChain
from langchain.chains.llm import LLMChain
from langchain.chains.question_answering import (
map_reduce_prompt,
refine_prompts,
stuff_prompt,
)
from langchain.chains.question_answering.map_rerank_prompt import (
PROMPT as MAP_RERANK_PROMPT,
)
[docs]class LoadingCallable(Protocol):
"""加载合并文档链的接口。"""
def __call__(
self, llm: BaseLanguageModel, **kwargs: Any
) -> BaseCombineDocumentsChain:
"""加载组合文档链的可调用函数。"""
def _load_map_rerank_chain(
llm: BaseLanguageModel,
prompt: BasePromptTemplate = MAP_RERANK_PROMPT,
verbose: bool = False,
document_variable_name: str = "context",
rank_key: str = "score",
answer_key: str = "answer",
callback_manager: Optional[BaseCallbackManager] = None,
callbacks: Callbacks = None,
**kwargs: Any,
) -> MapRerankDocumentsChain:
llm_chain = LLMChain(
llm=llm,
prompt=prompt,
verbose=verbose,
callback_manager=callback_manager,
callbacks=callbacks,
)
return MapRerankDocumentsChain(
llm_chain=llm_chain,
rank_key=rank_key,
answer_key=answer_key,
document_variable_name=document_variable_name,
verbose=verbose,
callback_manager=callback_manager,
**kwargs,
)
def _load_stuff_chain(
llm: BaseLanguageModel,
prompt: Optional[BasePromptTemplate] = None,
document_variable_name: str = "context",
verbose: Optional[bool] = None,
callback_manager: Optional[BaseCallbackManager] = None,
callbacks: Callbacks = None,
**kwargs: Any,
) -> StuffDocumentsChain:
_prompt = prompt or stuff_prompt.PROMPT_SELECTOR.get_prompt(llm)
llm_chain = LLMChain(
llm=llm,
prompt=_prompt,
verbose=verbose, # type: ignore[arg-type]
callback_manager=callback_manager,
callbacks=callbacks,
)
# TODO: document prompt
return StuffDocumentsChain(
llm_chain=llm_chain,
document_variable_name=document_variable_name,
verbose=verbose, # type: ignore[arg-type]
callback_manager=callback_manager,
callbacks=callbacks,
**kwargs,
)
def _load_map_reduce_chain(
llm: BaseLanguageModel,
question_prompt: Optional[BasePromptTemplate] = None,
combine_prompt: Optional[BasePromptTemplate] = None,
combine_document_variable_name: str = "summaries",
map_reduce_document_variable_name: str = "context",
collapse_prompt: Optional[BasePromptTemplate] = None,
reduce_llm: Optional[BaseLanguageModel] = None,
collapse_llm: Optional[BaseLanguageModel] = None,
verbose: Optional[bool] = None,
callback_manager: Optional[BaseCallbackManager] = None,
callbacks: Callbacks = None,
token_max: int = 3000,
**kwargs: Any,
) -> MapReduceDocumentsChain:
_question_prompt = (
question_prompt or map_reduce_prompt.QUESTION_PROMPT_SELECTOR.get_prompt(llm)
)
_combine_prompt = (
combine_prompt or map_reduce_prompt.COMBINE_PROMPT_SELECTOR.get_prompt(llm)
)
map_chain = LLMChain(
llm=llm,
prompt=_question_prompt,
verbose=verbose, # type: ignore[arg-type]
callback_manager=callback_manager,
callbacks=callbacks,
)
_reduce_llm = reduce_llm or llm
reduce_chain = LLMChain(
llm=_reduce_llm,
prompt=_combine_prompt,
verbose=verbose, # type: ignore[arg-type]
callback_manager=callback_manager,
callbacks=callbacks,
)
# TODO: document prompt
combine_documents_chain = StuffDocumentsChain(
llm_chain=reduce_chain,
document_variable_name=combine_document_variable_name,
verbose=verbose, # type: ignore[arg-type]
callback_manager=callback_manager,
callbacks=callbacks,
)
if collapse_prompt is None:
collapse_chain = None
if collapse_llm is not None:
raise ValueError(
"collapse_llm provided, but collapse_prompt was not: please "
"provide one or stop providing collapse_llm."
)
else:
_collapse_llm = collapse_llm or llm
collapse_chain = StuffDocumentsChain(
llm_chain=LLMChain(
llm=_collapse_llm,
prompt=collapse_prompt,
verbose=verbose, # type: ignore[arg-type]
callback_manager=callback_manager,
callbacks=callbacks,
),
document_variable_name=combine_document_variable_name,
verbose=verbose, # type: ignore[arg-type]
callback_manager=callback_manager,
)
reduce_documents_chain = ReduceDocumentsChain( # type: ignore[misc]
combine_documents_chain=combine_documents_chain,
collapse_documents_chain=collapse_chain,
token_max=token_max,
verbose=verbose,
)
return MapReduceDocumentsChain(
llm_chain=map_chain,
document_variable_name=map_reduce_document_variable_name,
reduce_documents_chain=reduce_documents_chain,
verbose=verbose, # type: ignore[arg-type]
callback_manager=callback_manager,
callbacks=callbacks,
**kwargs,
)
def _load_refine_chain(
llm: BaseLanguageModel,
question_prompt: Optional[BasePromptTemplate] = None,
refine_prompt: Optional[BasePromptTemplate] = None,
document_variable_name: str = "context_str",
initial_response_name: str = "existing_answer",
refine_llm: Optional[BaseLanguageModel] = None,
verbose: Optional[bool] = None,
callback_manager: Optional[BaseCallbackManager] = None,
callbacks: Callbacks = None,
**kwargs: Any,
) -> RefineDocumentsChain:
_question_prompt = (
question_prompt or refine_prompts.QUESTION_PROMPT_SELECTOR.get_prompt(llm)
)
_refine_prompt = refine_prompt or refine_prompts.REFINE_PROMPT_SELECTOR.get_prompt(
llm
)
initial_chain = LLMChain(
llm=llm,
prompt=_question_prompt,
verbose=verbose, # type: ignore[arg-type]
callback_manager=callback_manager,
callbacks=callbacks,
)
_refine_llm = refine_llm or llm
refine_chain = LLMChain(
llm=_refine_llm,
prompt=_refine_prompt,
verbose=verbose, # type: ignore[arg-type]
callback_manager=callback_manager,
callbacks=callbacks,
)
return RefineDocumentsChain(
initial_llm_chain=initial_chain,
refine_llm_chain=refine_chain,
document_variable_name=document_variable_name,
initial_response_name=initial_response_name,
verbose=verbose, # type: ignore[arg-type]
callback_manager=callback_manager,
callbacks=callbacks,
**kwargs,
)
[docs]def load_qa_chain(
llm: BaseLanguageModel,
chain_type: str = "stuff",
verbose: Optional[bool] = None,
callback_manager: Optional[BaseCallbackManager] = None,
**kwargs: Any,
) -> BaseCombineDocumentsChain:
"""加载问答链。
参数：
llm: 在链中使用的语言模型。
chain_type: 要使用的文档组合链的类型。应为"stuff"、"map_reduce"、"map_rerank"和"refine"之一。
verbose: 链是否应以详细模式运行。请注意，这适用于构成最终链的所有链。
callback_manager: 用于链的回调管理器。
返回：
用于问答的链。
"""
loader_mapping: Mapping[str, LoadingCallable] = {
"stuff": _load_stuff_chain,
"map_reduce": _load_map_reduce_chain,
"refine": _load_refine_chain,
"map_rerank": _load_map_rerank_chain,
}
if chain_type not in loader_mapping:
raise ValueError(
f"Got unsupported chain type: {chain_type}. "
f"Should be one of {loader_mapping.keys()}"
)
return loader_mapping[chain_type](
llm, verbose=verbose, callback_manager=callback_manager, **kwargs
)