Ng1yVUddlZddlZddlmZddlmZmZmZmZm Z m Z m Z m Z m Z mZddlZddlmZddlmZddlmZddlmZddlmZdd lmZdd lmZdd l m!Z! dd l"m#Z#m$Z$dd l%m&Z&n#e'$r dZ$dZ#dZ&YnwxYwe$rde$j(DniZ)ee*e e+ej,ffe-d<e#rde#j(DniZ.ee*e e+ej,ffe-d<e#rde#j(Dn e/Z0ee*e-d<GddeZ1dS)N)tee) AnyDictIterableListOptionalSequenceTupleUnionget_argsSet)grpc)AsyncQdrantBase) common_types) GrpcToRest)Document) QueryResponse)modelsreciprocal_rank_fusionSparseTextEmbedding TextEmbedding) OnnxProvidercNi|]"}|d|dtjjf#S)modeldim)rDistanceCOSINE.0rs `/var/www/html/ai-engine/env/lib/python3.11/site-packages/qdrant_client/async_qdrant_fastembed.py r#"s<  guv'=>SUPPORTED_EMBEDDING_MODELSc i|] }|d| S)rr s r"r#r#*sTTTuU7^UTTTr$!SUPPORTED_SPARSE_EMBEDDING_MODELScJh|] }|dd|d!S) requires_idfNr)get)r! model_configs r" r-/sC    ND 1 1Wr$IDF_EMBEDDING_MODELScDeZdZUdZiZeedfed<iZeedfed<e ed<de ffd Z e d efd Z e d eefd Z dAd edeedeedeedeedde d d fdZ dBd eedeedeedeedde d d f dZedCdZeded eeejffdZe dBdedeedeedeedde d df dZe dBdedeedeedeedde d df dZeddd fdeed edededeed eeeeeff d Zedd fdeed ededeed ee j!f d!Z"d efd"Z#d eefd#Z$d$ee j%d ee&fd%Z' dDd&eeej(d'eeeee fd(eeeeefd)e)d*eee j!d eej*f d+Z+d,ej,d d fd-Z- dBd.ee d/eej.d0eej/d eeej0ffd1Z1 dEd.ee d2eej2d eeeej3ffd3Z4 dFd4edeed'eeeee fd&eeej(dedeede d ee5eeffd5Z6 dDd6e5e j7eeeeee j!e j8e j9e:d fd7e5ej;eej;d fd eeej8eej;ffd8Zd=ede d ee&f d>Z? dGd4ed?eedd=ede d eee&f d@Z@xZAS)HAsyncQdrantFastembedMixinzBAAI/bge-small-enrembedding_modelsrsparse_embedding_models_FASTEMBED_INSTALLEDkwargsc Bd|_d|_ ddlm}m}t |jdksJt |jdksJd|j_n#t$rd|j_YnwxYwtj di|dS)NrrTFr') _embedding_model_name_sparse_embedding_model_name fastembedrrlenlist_supported_models __class__r3 ImportErrorsuper__init__)selfr4rrr;s r"r>z"AsyncQdrantFastembedMixin.__init__?s48";?) 8 D D D D D D D D@*@BBCCaGGGG:}:<<==AAAA26DN / / 8 8 827DN / / / 8""6"""""sAA**BBreturnc6|j |j|_|jSN)r6DEFAULT_EMBEDDING_MODELr?s r"embedding_model_namez.AsyncQdrantFastembedMixin.embedding_model_nameLs  % -)-)ED &))r$c|jSrB)r7rDs r"sparse_embedding_model_namez5AsyncQdrantFastembedMixin.sparse_embedding_model_nameRs 00r$NrE max_length cache_dirthreads providersrc t|tjdtd|jd||||d|||_dS)a Set embedding model to use for encoding documents and queries. Args: embedding_model_name: One of the supported embedding models. See `SUPPORTED_EMBEDDING_MODELS` for details. max_length (int, optional): Deprecated. Defaults to None. cache_dir (str, optional): The path to the cache directory. Can be set using the `FASTEMBED_CACHE_PATH` env variable. Defaults to `fastembed_cache` in the system's temp directory. threads (int, optional): The number of threads single onnxruntime session can use. Defaults to None. providers: The list of onnx providers (with or without options) to use. Defaults to None. Example configuration: https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html#configuration-options Raises: ValueError: If embedding model is not supported. ImportError: If fastembed is not installed. Returns: None Nzhmax_length parameter is deprecated and will be removed in the future. It's not used by fastembed models.) stacklevel model_namerIrJrKr')warningswarnDeprecationWarning_get_or_init_modelr6)r?rErHrIrJrKr4s r" set_modelz#AsyncQdrantFastembedMixin.set_modelVsu:  ! Mz"      +       &:"""r$c <||jd||||d|||_dS)ay Set sparse embedding model to use for hybrid search over documents in combination with dense embeddings. Args: embedding_model_name: One of the supported sparse embedding models. See `SUPPORTED_SPARSE_EMBEDDING_MODELS` for details. If None, sparse embeddings will not be used. cache_dir (str, optional): The path to the cache directory. Can be set using the `FASTEMBED_CACHE_PATH` env variable. Defaults to `fastembed_cache` in the system's temp directory. threads (int, optional): The number of threads single onnxruntime session can use. Defaults to None. providers: The list of onnx providers (with or without options) to use. Defaults to None. Example configuration: https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html#configuration-options Raises: ValueError: If embedding model is not supported. ImportError: If fastembed is not installed. Returns: None NrOr')_get_or_init_sparse_modelr7)r?rErIrJrKr4s r"set_sparse_modelz*AsyncQdrantFastembedMixin.set_sparse_modelsR8 + *D * /##        -A)))r$c2|jrdStd)Nzjfastembed is not installed. Please install it to enable fast vector indexing with `pip install fastembed`.)r3r<)clss r"_import_fastembedz+AsyncQdrantFastembedMixin._import_fastembeds'  #  F x   r$rPc||tvrtd|dtt|S)NUnsupported embedding model: . Supported models: )r[r% ValueError)rZrPs r"_get_model_paramsz+AsyncQdrantFastembedMixin._get_model_paramssS  7 7 7l llPjll **55r$c ||jvr |j|S||tvrtd|dtt d||||d||j|<|j|SNr]r^rOr')r1r[r%r_rrZrPrIrJrKr4s r"rTz,AsyncQdrantFastembedMixin._get_or_init_models - - -' 3 3  7 7 7l llPjll ,9, ! , ,  , , Z(#J//r$c ||jvr |j|S||tvrtd|dtt d||||d||j|<|j|Srb)r2r[r(r_rrcs r"rWz3AsyncQdrantFastembedMixin._get_or_init_sparse_models 4 4 4.z: :  > > >s ssPqss 3F3 ! 3 3  3 3 #J/*:66r$ default documents batch_size embed_typeparallelc# K|| t|d\}}|dkr |||}nF|dkr fd|D}n1|dkr |||}nt d|t ||D]\} } | | fVdS) NrPrMpassagerhrjqueryc3jK|]-}t|dV.dS)rorN)list query_embed)r!roembedding_models r" z=AsyncQdrantFastembedMixin._embed_documents..sRFK_00u0==>>qAr$rfzUnknown embed type: )rTr passage_embedembedr_ziptolist) r?rgrErhrirj documents_a documents_b vectors_itervectordocrts @r"_embed_documentsz*AsyncQdrantFastembedMixin._embed_documentss%11=Q1RR%(A%6%6"k  " "*88 X9LL7 " "OZLL9 $ $*00 X1LL@J@@AA A|[99 ) )KFC ( ( ( ( ( ) )r$c#K||}||||}|D]H}tj|j|jVIdS)Nrlrnindicesvalues)rWrwtypes SparseVectorrryr)r?rgrErhrjsparse_embedding_modelr| sparse_vectors r"_sparse_embed_documentsz1AsyncQdrantFastembedMixin._sparse_embed_documentss"&!?!?K_!?!`!`-33 *x4  *  M$%-4466}?S?Z?Z?\?\      r$cp|jdd}d|S) Returns name of the vector field in qdrant collection, used by current fastembed model. Returns: Name of the vector field. /zfast-)rEsplitlowerr?rPs r"get_vector_field_namez/AsyncQdrantFastembedMixin.get_vector_field_names9 .44S99"=CCEE #z###r$c|j7|jdd}d|SdS)rNrrz fast-sparse-)rGrrrs r"get_sparse_vector_field_namez6AsyncQdrantFastembedMixin.get_sparse_vector_field_namesG  + 79??DDRHNNPPJ.*.. .tr$ scored_pointsc g}|}|}|D]}t|jtr|j|dnd}d}|7t|jtr|j|dnd}|t|j|||j |j dd|j |S)Ndocument)id embeddingsparse_embeddingmetadatarscore) rr isinstancer}rr+appendrrpayloadr)r?rresponsevector_field_namesparse_vector_field_name scored_pointrrs r"!_scored_points_to_query_responsesz;AsyncQdrantFastembedMixin._scored_points_to_query_responses*s 6688#'#D#D#F#F )  Ll1488 #''(94@@@  $ '3","5t<<L'++,DdKKK! OO#'%5)1)155j"EE&,     r$idsr encoded_docsids_accumulatorsparse_vectorsc#K|tdd}|tdd}|tdd}|}|}t||||D]K\}} \} } } ||d| i| } || i}|| | ||<t j|| |VLdS)Nc2tjjSrB)uuiduuid4hexr'r$r"z.Rstz||/r$ciSrBr'r'r$r"rz.TsBr$cdSrBr'r'r$r"rz.Vs$r$Tr)rrr})iterrrrxrr PointStruct)r?rrrrr vector_namesparse_vector_nameidxmetar~r}rr point_vectors r"_points_iteratorz*AsyncQdrantFastembedMixin._points_iteratorIs ;//66C  JJ--H  !!,,55N0022 !>>@@7: <8 8  S S 3C}Vm  " "3 ' ' '!3/$/G6A65JL!--2K3@ /0$W\RRR R R R R S Sr$collection_infoc4||j\}}|}t|jjjtsJd|jjj||jjjvsJd|jjjd||jjj|}||jksJd|d|j||j ksJd|d|j | }|||jjj vsJd|jjj|j tvrG|jjj |j}|tjjksJ|j d|dSdSdS)Nrlz,Collection have incompatible vector params: z , expected zEmbedding size mismatch: z != zDistance mismatch: z, requires modifier IDF, current modifier is )r`rErrconfigparamsvectorsdictsizedistancerrrGr.modifierrModifierIDF)r?rembeddings_sizerr vector_paramsrrs r"_validate_collection_infoz3AsyncQdrantFastembedMixin._validate_collection_infocs&*&<&<Ha&<&b&b#( 6688  " ) 14   b b a/:P:W:_ a a b b b !7!>!F F F F /:P:W:_  l}   G F F'.5=>OP }1 1 1 1 P P PMT>[>ceeYXX/3GGG*18G, 33336nndlnn433 0 /HG 43r$on_diskquantization_config hnsw_configc|}||j\}}|tj|||||iS)a Generates vector configuration, compatible with fastembed models. Args: on_disk: if True, vectors will be stored on disk. If None, default value will be used. quantization_config: Quantization configuration. If None, quantization will be disabled. hnsw_config: HNSW configuration. If None, default configuration will be used. Returns: Configuration for `vectors_config` argument in `create_collection` method. rl)rrrrr)rr`rEr VectorParams)r?rrrrrrs r"get_fastembed_vector_paramsz5AsyncQdrantFastembedMixin.get_fastembed_vector_paramssd"!6688&*&<&<Ha&<&b&b#( v2$!$7'      r$rc|}|jtvr|tjjn|}|dS|tjtj||iS)a Generates vector configuration, compatible with fastembed sparse models. Args: on_disk: if True, vectors will be stored on disk. If None, default value will be used. modifier: Sparse vector queries modifier. E.g. Modifier.IDF for idf-based rescoring. Default: None. Returns: Configuration for `vectors_config` argument in `create_collection` method. N)r)indexr)rrGr.rrrSparseVectorParamsSparseIndexParams)r?rrrs r""get_fastembed_sparse_vector_paramszv**HH  $4 v8.w???(     r$collection_namec :K|||j|d|}d} |j|||j||} ||d{V} nn#t $ra||||d{V||d{V} YnwxYw| | g} | |||| | } |j d || d|pd |d || S) a Adds text documents into qdrant collection. If collection does not exist, it will be created with default parameters. Metadata in combination with documents will be added as payload. Documents will be embedded using the specified embedding model. If you want to use your own vectors, use `upsert` method instead. Args: collection_name (str): Name of the collection to add documents to. documents (Iterable[str]): List of documents to embed and add to the collection. metadata (Iterable[Dict[str, Any]], optional): List of metadata dicts. Defaults to None. ids (Iterable[models.ExtendedPointId], optional): List of ids to assign to documents. If not specified, UUIDs will be generated. Defaults to None. batch_size (int, optional): How many documents to embed and upload in single request. Defaults to 32. parallel (Optional[int], optional): How many parallel workers to use for embedding. Defaults to None. If number is specified, data-parallel process will be used. Raises: ImportError: If fastembed is not installed. Returns: List of IDs of added documents. If no ids provided, UUIDs will be randomly generated on client side. rm)rgrErhrirjN)rgrErhrj)r)rvectors_configsparse_vectors_config)rrrrrT)rpointswaitrjrhr') rrErGrget_collection Exceptioncreate_collectionrrrr upload_points) r?rrgrrrhrjr4rencoded_sparse_docsr inserted_idsrs r"addzAsyncQdrantFastembedMixin.addsR,,!%!:!  -  #  + 7"&">">#%)%E%! #?##   Y$($7$7$7$X$XXXXXXXOO Y Y Y(( /#??AA&*&M&M&O&O)        %)$7$7$7$X$XXXXXXXOOO  Y &&777 &&%(. '    +]!       s A''A(CCroprefetchcn||}|g}t|ts|g}||fS)Nrq)&_resolve_query_to_embedding_embeddingsrrr)r?rors r"3_resolve_query_to_embedding_embeddings_and_prefetchzMAsyncQdrantFastembedMixin._resolve_query_to_embedding_embeddings_and_prefetchsI;;%;HH  H(D)) " zHx  r$ct|ttjst|tjr|St|tjrt j|St|tj r't j| St|trt j|St|ttj rEt|tj rtj|n|}t j|S|dSt|tr]|j}|t#d|t$vrm||||}t||jgd }n|t.vr||||}t||jgd}t j|j |j }nt#|dt j|St#dt9|) N)nearestzH`query_points` requires explicit model name specification for `Document`rlrgrrz is not among supported modelszUnsupported query type: )rr rQueryrrr NearestQuerynpndarrayryrrPointIdrconvert_point_idrrr_r%rUrTrwtextr(rXrWrrtype)r?rorPembedding_model_instrsparse_embedding_model_insts r"rz@AsyncQdrantFastembedMixin._resolve_query_to_embedding_embeddingss eXek22 3 3 z%7T7T L eU/ 0 0 6&u555 5 eRZ ( ( ?&u||~~>>> > eT " " 6&u555 5 eXem44 5 5 66@ 6U6U` +E222[` &u555 5 =4 eX & & :J! ^777z***'+'>'>*'>'U'U$ !5!;!;uzl!;!S!STTUVW^^`` @@@%%j111.2.L.LXb.L.c.c+ !z9AsyncQdrantFastembedMixin.query_batch..s'___D::8DD___r$cg|]F}tj|j|jGS)r)rrrryr)r!rs r"rz9AsyncQdrantFastembedMixin.query_batch..s_     %-4466}?S?Z?Z?\?\      r$rc<g|]\}}t||gS)rr)r!dense_responsesparse_responsers r"rz9AsyncQdrantFastembedMixin.query_batch..s?   1 #NO#DE R R R   r$c:g|]}|Sr'r r s r"rz9AsyncQdrantFastembedMixin.query_batch..s'[[[X66x@@[[[r$r')rTrErrrsrrrrryrrGrrWrwrrr9rx)r?rr rrr4r query_vectorsrr}request responsesrsparse_query_vectorsrdense_responsessparse_responsess` ` r" query_batchz%AsyncQdrantFastembedMixin.query_batchs8 $66$B[6\\1==K=PPQQ # % %F*)3355fmmoo$! G OOG $ $ $ $  + 3"//Zb/ccccccccI____U^___ _&*&D&D7'E' ' #  "=!B!B[!B!Y!Y     2 % %M*/::<<]$! G OOG $ $ $ $++OV^+________ #$6c+&6&6$67$S%5%5%7%78    58JZ5[5[   \[[[QZ[[[[r$)NNNN)NNN)r@NrB)NN)NNreN)Nr)B__name__ __module__ __qualname__rCr1rstr__annotations__r2boolrr>propertyrErrGintr rUrX classmethodr[r rrr`rTrWrrfloatrrrrrr ScoredPointrrExtendedPointIdrrrrCollectionInforQuantizationConfigHnswConfigDiffrrrrrr rrr NumpyArrayrPrefetchrrFilterror __classcell__)r;s@r"r0r09s135d3/0555@BT#'<"<=BBB # # # # # # #*c***X* 1Xc]111X1 %)#'!%8< *:*:!*:SM*:C= *: # *: H^45 *:*: *:*:*:*:^$(!%8< $A$A&sm$AC=$A# $A H^45 $A  $A $A$A$A$AL   [ 6365fo9M3N666[6$(!%8< 000C=0# 0 H^45 0  0 000[00$(!%8< 777C=7# 7 H^45 7  7 777[76%<#"& ))C=)") )  ) 3- ) %T%[() *))))<%<"& C="  3-  %$ %  $s$$$$ hsm    !%"34 m JBF SS hv56 7S8DcN34SuS$u+%567 S  S !%* S &$ %SSSS4o9NoSWoooo>#'CG7;   $ &f&?@ f34  c6&& '     <UY  ~ 8@8Q $sF556 7    68<:>"&RRRC=R8DcN34 R hv56 7 R  R3-RR eCHo RRRR@IM!! M K e    K       !fo)>DE! x %tFO'<< =!!!!*0C M K e    K       0C &, 0C0C0C0Cl15 H H H H v}- H  H  H  m H H H H \15 H\H\H\#YH\v}- H\  H\  H\ d=! "H\H\H\H\H\H\H\H\r$r0)2rrQ itertoolsrtypingrrrrrr r r r r numpyr qdrant_clientrqdrant_client.async_client_baserqdrant_client.conversionsrr$qdrant_client.conversions.conversionrqdrant_client.embed.modelsrqdrant_client.fastembed_commonrqdrant_client.httprqdrant_client.hybrid.fusionrr8rrfastembed.commonrr<r:r%rr"rrr(setr.r0r'r$r"r;s ]]]]]]]]]]]]]]]]]]]]]]]];;;;;;;;;;;;;;;;;;//////888888%%%%%%>>>>>><<<<<<<<-------MLLL  8]8:: DeC,@&A!AB TT(Q(;(Q(S(STTTT "4U33G-H(H#I E/EGG c#hh \h \h \h \h \h \h \h \h \h \sA,, A:9A: