langchain-aws

This package contains the LangChain integrations with AWS.

Installation

pip install -U langchain-aws

All integrations in this package assume that you have the credentials setup to connect with AWS services.

Authentication

In order to use Amazon Bedrock models, you need to configure AWS credentials. One of the options is to set the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY environment variables. More information can be found here. Alternatively, set the AWS_BEARER_TOKEN_BEDROCK environment variable locally for API Key authentication. For additional API key details, refer to docs.

Chat Models

ChatBedrock class exposes chat models from Bedrock.

from langchain_aws import ChatBedrock

llm = ChatBedrock()
llm.invoke("Sing a ballad of LangChain.")

Embeddings

BedrockEmbeddings class exposes embeddings from Bedrock.

from langchain_aws import BedrockEmbeddings

embeddings = BedrockEmbeddings()
embeddings.embed_query("What is the meaning of life?")

LLMs

BedrockLLM class exposes LLMs from Bedrock.

from langchain_aws import BedrockLLM

llm = BedrockLLM()
llm.invoke("The meaning of life is")

Retrievers

AmazonKendraRetriever class provides a retriever to connect with Amazon Kendra.

from langchain_aws import AmazonKendraRetriever

retriever = AmazonKendraRetriever(
    index_id="561be2b6d-9804c7e7-f6a0fbb8-5ccd350"
)

retriever.get_relevant_documents(query="What is the meaning of life?")

AmazonKnowledgeBasesRetriever class provides a retriever to connect with Amazon Knowledge Bases.

from langchain_aws import AmazonKnowledgeBasesRetriever

retriever = AmazonKnowledgeBasesRetriever(
    knowledge_base_id="IAPJ4QPUEU",
    retrieval_config={"vectorSearchConfiguration": {"numberOfResults": 4}},
)

retriever.get_relevant_documents(query="What is the meaning of life?")

VectorStores

InMemoryVectorStore class provides a vectorstore to connect with Amazon MemoryDB.

from langchain_aws.vectorstores.inmemorydb import InMemoryVectorStore

vds = InMemoryVectorStore.from_documents(
            chunks,
            embeddings,
            redis_url="rediss://cluster_endpoint:6379/ssl=True ssl_cert_reqs=none",
            vector_schema=vector_schema,
            index_name=INDEX_NAME,
        )

MemoryDB as Retriever

Here we go over different options for using the vector store as a retriever.

There are three different search methods we can use to do retrieval. By default, it will use semantic similarity.

retriever=vds.as_retriever()