LLM LangChain

使用浪链

LangChain是一个帮助您方便使用LM（Large Language）的框架。

基本的

LangChain Basic 解释了 LangChain 每种配置的示例代码。

在Falcon FM上使用LangChain

描述如何将 LangChain 应用到使用 Falcon FM 创建的 SageMaker 端点。使用通过 SageMaker JumpStart 安装 Falcon FM 获得的 SageMaker 端点（例如，jumpstart-dft-hf-llm-falcon-7b-instruct-bf16）。

SageMaker Endpoint 的 LangChain 声明

参考Falcon的输入和输出，注册ContentHandler的transform_input和transform_output，如下所示。

 from langchain import PromptTemplate , SagemakerEndpoint
from langchain . llms . sagemaker_endpoint import LLMContentHandler

class ContentHandler ( LLMContentHandler ):
    content_type = "application/json"
    accepts = "application/json"

    def transform_input ( self , prompt : str , model_kwargs : dict ) -> bytes :
        input_str = json . dumps ({ 'inputs' : prompt , 'parameters' : model_kwargs })
        return input_str . encode ( 'utf-8' )
      
    def transform_output ( self , output : bytes ) -> str :
        response_json = json . loads ( output . read (). decode ( "utf-8" ))        
        return response_json [ 0 ][ "generated_text" ]

使用 endpoint_name、aws_region、parameters 和 content_handler 为 Sagemaker Endpoint 注册 llm，如下所示。

 endpoint_name = 'jumpstart-dft-hf-llm-falcon-7b-instruct-bf16'
aws_region = boto3 . Session (). region_name
parameters = {
    "max_new_tokens" : 300
}
content_handler = ContentHandler ()

llm = SagemakerEndpoint (
    endpoint_name = endpoint_name , 
    region_name = aws_region , 
    model_kwargs = parameters ,
    content_handler = content_handler
)

您可以按如下方式检查 llm 的运行情况。

 llm ( "Tell me a joke" )

此时的结果如下。

 I once told a joke to a friend, but it didn't work. He just looked

网页加载器

Web 加载器 - 您可以使用 langchain 加载网页。

 from langchain . document_loaders import WebBaseLoader
from langchain . indexes import VectorstoreIndexCreator

loader = WebBaseLoader ( "https://lilianweng.github.io/posts/2023-06-23-agent/" )
index = VectorstoreIndexCreator (). from_loaders ([ loader ])

提示模板

定义如下所示的模板后，您可以定义LLMChain并运行它。有关详细信息，请参阅 langchain-sagemaker-endpoint-Q&A.ipynb。

 from langchain import PromptTemplate , LLMChain

template = "Tell me a {adjective} joke about {content}."
prompt = PromptTemplate . from_template ( template )

llm_chain = LLMChain ( prompt = prompt , llm = llm )

outputText = llm_chain . run ( adjective = "funny" , content = "chickens" )
print ( outputText )

此时的结果如下。

 Why did the chicken cross the playground? To get to the other slide!

问题/回答

使用 langchain.chains.question_answering 执行文档的提问/回答。有关详细信息，请参阅 langchain-sagemaker-endpoint-Q&A.ipynb。

定义提示的模板。

 template = """Use the following pieces of context to answer the question at the end.

{context}

Question: {question}
Answer:"""

prompt = PromptTemplate (
    template = template , input_variables = [ "context" , "question" ]
)

使用 langchain.docstore.document 创建文档。

 from langchain . docstore . document import Document
example_doc_1 = """
Peter and Elizabeth took a taxi to attend the night party in the city. While in the party, Elizabeth collapsed and was rushed to the hospital.
Since she was diagnosed with a brain injury, the doctor told Peter to stay besides her until she gets well.
Therefore, Peter stayed with her at the hospital for 3 days without leaving.
"""

docs = [
    Document (
        page_content = example_doc_1 ,
    )
]

现在进行提问/回答。

 from langchain . chains . question_answering import load_qa_chain

question = "How long was Elizabeth hospitalized?"

chain = load_qa_chain ( prompt = prompt , llm = llm )

output = chain ({ "input_documents" : docs , "question" : question }, return_only_outputs = True )
print ( output )

此时的结果如下。

 {'output_text': ' 3 days'}

PDF摘要

langchain-sagemaker-endpoint-pdf-summary.ipynb 解释了如何使用基于 Falcon FM 的 SageMaker Endpoint 进行 PDF Summery。

首先，使用 PyPDF2 读取存储在 S3 中的 PDF 文件并提取文本。

 import PyPDF2
from io import BytesIO

sess = sagemaker . Session ()
s3_bucket = sess . default_bucket ()
s3_prefix = 'docs'
s3_file_name = '2016-3series.pdf'   # S3의 파일명


s3r = boto3 . resource ( "s3" )
doc = s3r . Object ( s3_bucket , s3_prefix + '/' + s3_file_name )
        
contents = doc . get ()[ 'Body' ]. read ()
reader = PyPDF2 . PdfReader ( BytesIO ( contents ))
        
raw_text = []
for page in reader . pages :
    raw_text . append ( page . extract_text ())
contents = ' n ' . join ( raw_text )

new_contents = str ( contents ). replace ( " n " , " " )

由于文档很大，使用RecursiveCharacterTextSplitter将其分成块并保存在Document中。然后，使用load_summarize_chain进行汇总。

 from langchain . text_splitter import CharacterTextSplitter
from langchain . text_splitter import RecursiveCharacterTextSplitter

text_splitter = RecursiveCharacterTextSplitter ( chunk_size = 1000 , chunk_overlap = 0 )
texts = text_splitter . split_text ( new_contents ) 

from langchain . docstore . document import Document
docs = [
    Document (
        page_content = t
    ) for t in texts [: 3 ]
]

from langchain . chains . summarize import load_summarize_chain
from langchain . prompts import PromptTemplate

prompt_template = """Write a concise summary of the following:


{text}


CONCISE SUMMARY """

PROMPT = PromptTemplate ( template = prompt_template , input_variables = [ "text" ])
chain = load_summarize_chain ( llm , chain_type = "stuff" , prompt = PROMPT )
summary = chain . run ( docs )

链类型

• stuff 将所有块放入一个提示中。因此，这将达到代币的最大限制。
• map_reduce 对每个块进行汇总、合并汇总、汇总合并汇总。如果合并的摘要太大，则会引发错误。
• 精炼总结第一个块，然后用第一个总结总结第二个块。重复相同的过程，直到汇总所有块。

ETC

浪链 by Bedrock

 from langchain import Bedrock
from langchain . embeddings import BedrockEmbeddings

llm = Bedrock ()

print ( llm ( "explain GenAI" ))

参考

浪链文档

浪链-github

SageMaker 端点

2-Lab02-RAG-法学硕士

AWS Kendra Langchain 扩展

质量检查和文档聊天

LangChain - 模块 - 语言模型 - 法学硕士 - 集成 - SageMakerEndpoint

LangChain - 生态系统 - 集成 - SageMaker Endpoint

摄取知识库数据 ta Vector DB