o Zh@sddlmZmZmZmZddlmZddlmZddl m Z m Z ddl m Z Gddde eeeeefZGdd d e eZd S) )AnyDictListOptional) Embeddings)pre_init) BaseModel ConfigDict)ContentHandlerBasec@seZdZdZdS)EmbeddingsContentHandlerzContent handler for LLM class.N)__name__ __module__ __qualname____doc__rrh/var/www/html/lang_env/lib/python3.10/site-packages/langchain_community/embeddings/sagemaker_endpoint.pyr sr c @seZdZUdZ dZeed<dZeed< dZ eed< dZ e eed< e ed< dZ e eed < dZe eed < ed d d dZededefddZdeedeeefddZ ddeededeeefddZdedeefddZdS)SagemakerEndpointEmbeddingsaCustom Sagemaker Inference Endpoints. To use, you must supply the endpoint name from your deployed Sagemaker model & the region where it is deployed. To authenticate, the AWS client uses the following methods to automatically load credentials: https://boto3.amazonaws.com/v1/documentation/api/latest/guide/credentials.html If a specific credential profile should be used, you must pass the name of the profile from the ~/.aws/credentials file that is to be used. Make sure the credentials / roles used have the required policies to access the Sagemaker endpoint. See: https://docs.aws.amazon.com/IAM/latest/UserGuide/access_policies.html Nclient endpoint_name region_namecredentials_profile_namecontent_handler model_kwargsendpoint_kwargsTZforbidr)Zarbitrary_types_allowedextraZprotected_namespacesvaluesreturnc Cs|ddur |S z;ddl}z"|ddur|j|dd}n|}|jd|dd|d<WW|StyE}ztd ||d}~wwtyPtd w) z.Dont do anything if client provided externallyrNrr)Z profile_namezsagemaker-runtimer)rzCould not load credentials to authenticate with AWS client. Please check that credentials in the specified profile name are valid. zRCould not import boto3 python package. Please install it with `pip install boto3`.)getboto3Sessionr Exception ValueError ImportError)clsrrsessionerrrvalidate_environmentus<  z0SagemakerEndpointEmbeddings.validate_environmenttextsc Csttdd|}|jp i}|jpi}|j||}|jj}|jj}z|jj d|j |||d|}Wnt yF}zt d|d}~ww|j |dS)z3Call out to SageMaker Inference embedding endpoint.cSs |ddS)N  )replace)xrrrs z=SagemakerEndpointEmbeddings._embedding_func..)Z EndpointNameBodyZ ContentTypeAcceptz$Error raised by inference endpoint: Nr.r)listmaprrrZtransform_input content_typeacceptsrZinvoke_endpointrr!r"Ztransform_output) selfr(Z _model_kwargsZ_endpoint_kwargsbodyr2r3responser&rrr_embedding_funcs(   z+SagemakerEndpointEmbeddings._embedding_func@ chunk_sizecCsVg}|t|kr t|n|}tdt||D]}|||||}||q|S)aCompute doc embeddings using a SageMaker Inference Endpoint. Args: texts: The list of texts to embed. chunk_size: The chunk size defines how many input texts will be grouped together as request. If None, will use the chunk size specified by the class. Returns: List of embeddings, one for each text. r)lenranger7extend)r4r(r9results _chunk_sizeir6rrrembed_documentss  z+SagemakerEndpointEmbeddings.embed_documentstextcCs||gdS)zCompute query embeddings using a SageMaker inference endpoint. Args: text: The text to embed. Returns: Embeddings for the text. r)r7)r4rArrr embed_querys z'SagemakerEndpointEmbeddings.embed_query)r8)r r rrrr__annotations__rstrrrrr rrrr Z model_configrr'rfloatr7intr@rBrrrrrs@    $  rN)typingrrrrZlangchain_core.embeddingsrZlangchain_core.utilsrZpydanticrr Z+langchain_community.llms.sagemaker_endpointr rDrEr rrrrrs   $