jpddlmZmZmZmZmZmZddlZddlm Z ddl m Z ddl m Z mZmZGdde e Zy))AnyDictListMappingOptionalTupleN) Embeddingsget_from_dict_or_env) BaseModel ConfigDictmodel_validatorc 6eZdZUdZdZeed< dZeed< dZeed< dZ e ed < d Z e eed <e d ZedededefdZedeeeffdZ ddeeeefdedeee fdZdeedeee fdZdedee fdZy )MosaicMLInstructorEmbeddingsaaMosaicML embedding service. To use, you should have the environment variable ``MOSAICML_API_TOKEN`` set with your API token, or pass it as a named parameter to the constructor. Example: .. code-block:: python from langchain_community.llms import MosaicMLInstructorEmbeddings endpoint_url = ( "https://models.hosted-on.mosaicml.hosting/instructor-large/v1/predict" ) mosaic_llm = MosaicMLInstructorEmbeddings( endpoint_url=endpoint_url, mosaicml_api_token="my-api-key" ) zBhttps://models.hosted-on.mosaicml.hosting/instructor-xl/v1/predict endpoint_urlz&Represent the document for retrieval: embed_instructionz (:#$ cd|jiS)zGet the identifying parameters.r)r)selfs r_identifying_paramsz0MosaicMLInstructorEmbeddings._identifying_params:s 1 122r!inputis_retryc d|i}|jdd} tj|j||} |jdk(rL|s2ddl}|j|j|j|d St d |j|j}t|trQgd } | D] } | |vs|| } nt d |t| t rt| dt r| } | S| g} | St d |#tjj $r}t d|d}~wwxYw#tjj"$r }t d|d|jd}~wwxYw)Ninputszapplication/json) Authorizationz Content-Type)headersjsonz$Error raised by inference endpoint: irT)r&z>Error raised by inference API: rate limit exceeded. Response: )dataoutputoutputsz#No key data or output in response: zUnexpected response type: zError raised by inference API: z . Response: )rrequestspostr exceptionsRequestException ValueError status_codetimesleepr_embedtextr+ isinstancedictlistJSONDecodeError) r#r%r&payloadr*responseer5parsed_response output_keyskey output_item embeddingss rr7z#MosaicMLInstructorEmbeddings._embed?sU#!% 7 78.   I}}T%6%6gVH' ##s*JJt//0;;ut;<< U}}o' 'mmoO/40; &Co-&5c&: ' %=o=NOk40Z APT5U!,J#.J!#=o=N!OPPK""33 ICA3GH H IN""22 1!M(--Q  sH"DAE>EF#E>>Ftextsch|Dcgc]}|j|f}}|j|}|Scc}w)zEmbed documents using a MosaicML deployed instructor embedding model. Args: texts: The list of texts to embed. Returns: List of embeddings, one for each text. )rr7)r#rEr8instruction_pairsrDs rembed_documentsz,MosaicMLInstructorEmbeddings.embed_documents{sAINNd44d;N[[!23 Os/r8cL|j|f}|j|gd}|S)zEmbed a query using a MosaicML deployed instructor embedding model. Args: text: The text to embed. Returns: Embeddings for the text. r)rr7)r#r8instruction_pair embeddings r embed_queryz(MosaicMLInstructorEmbeddings.embed_querys2!22D9KK!1 23A6 r!)F)__name__ __module__ __qualname____doc__rstr__annotations__rrrfloatrrr model_configr classmethodrrr propertyrr$rrboolr7rHrLr!rrr s-( M#EsE.Fs/KE(, ,L(#$3$3WS#X%633 >C:%S/*:6:: d5k :x T#Y 4U 3D   U r!r)typingrrrrrrr/langchain_core.embeddingsr langchain_core.utilsr pydanticr r rrrXr!rr]s*<<05;;J9jJr!