NgoddlmZddlZddlZddlZddlZddlmZddlm Z ddl m Z m Z m Z mZmZmZmZmZmZmZmZmZmZddlZddlmZddlmZddlmZdd l m!Z!dd l"m#Z#dd l$m%Z%m&Z&dd l'm(Z(dd l)m*Z*ddl+m,Z,e r$ee-ee-e.e/e0e1ffZ2ee2e(j3fZ4Gdde5Z6ddZ7edddGdde#Z8dS)) annotationsN)islice) itemgetter) TYPE_CHECKINGAnyAsyncGeneratorCallableDict GeneratorIterableListOptionalSequenceTupleTypeUnion) deprecated)Document) Embeddings)run_in_executor) VectorStore)AsyncQdrantClient QdrantClient)models)AsyncQdrantLocal)maximal_marginal_relevanceceZdZdZdS)QdrantExceptionz`Qdrant` related exceptions.N)__name__ __module__ __qualname____doc__Y/var/www/html/ai-engine/env/lib/python3.11/site-packages/langchain_qdrant/vectorstores.pyrr*s&&&&r$rmethodr returncHtjdfd }|S)z Decorator to call the synchronous method of the class if the async method is not implemented. This decorator might be only used for the methods that are defined as async in the class. selfrargskwargsr'c K |g|Ri|d{VS#t$r5tdt|jddg|Ri|d{VcYSwxYw)N)NotImplementedErrorrgetattrr)r)r*r+r&s r%wrapperz#sync_call_fallback..wrapper5s 6t666v66666666 6"    )gdFOABB$788;?CI     sZ ddd?Ze ddd@Z dddFZe dddGZ dddHZe dddIZ dddJZe dddKZdddNZe dddOZedddddPdQdRddddddddSeeeddddddddddddRfdduZedddddPdQdRdddddeed efddvZ ee dddddPdQdRddddddddSeeeddddddddddddRfddwZ!edddPdQdRddddddddSeeedddddddddddRfddxZ"edddPdQdRddddddddSeeedddddddddddRfddyZ#e$dd{Z%dd}Z& ddd~Z'e dddZ(eddZ)eddZ*ddZ+ddZ,ddZ-ddZ.ddZ/ddZ0 dddZ1 dddZ2e$ dddZ3dS)QdrantaG`Qdrant` vector store. Example: .. code-block:: python from qdrant_client import QdrantClient from langchain_qdrant import Qdrant client = QdrantClient() collection_name = "MyCollection" qdrant = Qdrant(client, collection_name, embedding_function) page_contentstr CONTENT_KEYmetadata METADATA_KEYN Optional[str] VECTOR_NAMECOSINEclientrcollection_name embeddingsOptional[Embeddings]content_payload_keymetadata_payload_keydistance_strategy vector_name async_client Optional[Any]embedding_functionOptional[Callable]c t|tstdt||4t|tstdt||| td|| td||_| |_||_||_||_ |p|j |_ |p|j |_ |p|j|_| t!jdt|t$s"t!jd||_d|_||_dS)z%Initialize with necessary components.z@client should be an instance of qdrant_client.QdrantClient, got NzIasync_client should be an instance of qdrant_client.AsyncQdrantClientgot z=`embeddings` value can't be None. Pass `Embeddings` instance.zMBoth `embeddings` and `embedding_function` are passed. Use `embeddings` only.z]Using `embedding_function` is deprecated. Pass `Embeddings` instance to `embeddings` instead.zq`embeddings` should be an instance of `Embeddings`.Using `embeddings` as `embedding_function` which is deprecated) isinstancer ValueErrortyper _embeddings_embeddings_functionrBrJrCr<rFr>rGr@rIwarningswarnrupperrH) r)rBrCrDrFrGrHrIrJrLs r%__init__zQdrant.__init__Xs&,// &F||&&   #J|EV,W,W #,L)),,   "4"<O   !&8&D)  &$6!$* 9E.#6#J$:J $8$MD$=    '+#      C 1 w//000r$c K|j||f|||||d|d{V} tttd| S)aReturn docs most similar to embedding vector. Args: embedding: Embedding vector to look up documents similar to. k: Number of Documents to return. Defaults to 4. filter: Filter by metadata. Defaults to None. search_params: Additional search params offset: Offset of the first result to return. May be used to paginate results. Note: large offset values may cause performance issues. score_threshold: Define a minimal score threshold for the result. If defined, less similar results will not be returned. Score of the returned result might be higher or smaller than the threshold depending on the Distance function used. E.g. for cosine similarity only higher scores will be returned. consistency: Read consistency of the search. Defines how many replicas should be queried before returning the result. Values: - int - number of replicas to query, values should present in all queried replicas - 'majority' - query all replicas, but return values present in the majority of replicas - 'quorum' - query the majority of replicas, return values present in all of them - 'all' - query all replicas, and return values present in all replicas **kwargs: Any other named arguments to pass through to AsyncQdrantClient.Search(). Returns: List of Documents most similar to the query. rNr)rrrrrs r%asimilarity_search_by_vectorz#Qdrant.asimilarity_search_by_vectors^ED    '+#            C 1 w//000r$c |Et|tr0tjdt|} n|} |} j j|f} jjdj | | |||dd||d |} fd| DS)aReturn docs most similar to embedding vector. Args: embedding: Embedding vector to look up documents similar to. k: Number of Documents to return. Defaults to 4. filter: Filter by metadata. Defaults to None. search_params: Additional search params offset: Offset of the first result to return. May be used to paginate results. Note: large offset values may cause performance issues. score_threshold: Define a minimal score threshold for the result. If defined, less similar results will not be returned. Score of the returned result might be higher or smaller than the threshold depending on the Distance function used. E.g. for cosine similarity only higher scores will be returned. consistency: Read consistency of the search. Defines how many replicas should be queried before returning the result. Values: - int - number of replicas to query, values should present in all queried replicas - 'majority' - query all replicas, but return values present in the majority of replicas - 'quorum' - query the majority of replicas, return values present in all of them - 'all' - query all replicas, and return values present in all replicas **kwargs: Any other named arguments to pass through to QdrantClient.search() Returns: List of documents most similar to the query text and distance for each. NUsing dict as a `filter` is deprecated. Please use qdrant-client filters directly: https://qdrant.tech/documentation/concepts/filtering/TF rC query_vector query_filterrzlimitr| with_payload with_vectorsr}rclg|]0}|jjj|jf1Sr#_document_from_scored_pointrCrFrGscore.0resultr)s r% zAQdrant.similarity_search_with_score_by_vector..a\    00(,-        r$r#) rOdictrTrUDeprecationWarning_qdrant_filter_from_dictrIrBsearchrC r)rrwrxrzr|r}rr+ qdrant_filterrrs ` r%rz-Qdrant.similarity_search_with_score_by_vectorsZ  *VT":":  MH#     !99&AAMM"M   ' ,i8L$$+$   0%&'+#           "    r$cKjtjjtrt d|Et|t r0t jdt |} n|} |} j j |f} jj dj | | |||dd||d |d{V} fd| DS) aReturn docs most similar to embedding vector. Args: embedding: Embedding vector to look up documents similar to. k: Number of Documents to return. Defaults to 4. filter: Filter by metadata. Defaults to None. search_params: Additional search params offset: Offset of the first result to return. May be used to paginate results. Note: large offset values may cause performance issues. score_threshold: Define a minimal score threshold for the result. If defined, less similar results will not be returned. Score of the returned result might be higher or smaller than the threshold depending on the Distance function used. E.g. for cosine similarity only higher scores will be returned. consistency: Read consistency of the search. Defines how many replicas should be queried before returning the result. Values: - int - number of replicas to query, values should present in all queried replicas - 'majority' - query all replicas, but return values present in the majority of replicas - 'quorum' - query the majority of replicas, return values present in all of them - 'all' - query all replicas, and return values present in all replicas **kwargs: Any other named arguments to pass through to AsyncQdrantClient.Search(). Returns: List of documents most similar to the query text and distance for each. NrqrTFrclg|]0}|jjj|jf1Sr#rrs r%rzBQdrant.asimilarity_search_with_score_by_vector..rr$r#) rJrOrrrr.rrTrUrrrIrrCrs ` r%rz.Qdrant.asimilarity_search_with_score_by_vectornsI`   $   %'7) ) $&M   *VT":":  MH#     !99&AAMM"M   ' ,i8L0)0   0%&'+#                 "    r$?fetch_k lambda_multfloatc X||} |j| f|||||||d| S)aReturn docs selected using the maximal marginal relevance. Maximal marginal relevance optimizes for similarity to query AND diversity among selected documents. Args: query: Text to look up documents similar to. k: Number of Documents to return. Defaults to 4. fetch_k: Number of Documents to fetch to pass to MMR algorithm. Defaults to 20. lambda_mult: Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5. filter: Filter by metadata. Defaults to None. search_params: Additional search params score_threshold: Define a minimal score threshold for the result. If defined, less similar results will not be returned. Score of the returned result might be higher or smaller than the threshold depending on the Distance function used. E.g. for cosine similarity only higher scores will be returned. consistency: Read consistency of the search. Defines how many replicas should be queried before returning the result. Values: - int - number of replicas to query, values should present in all queried replicas - 'majority' - query all replicas, but return values present in the majority of replicas - 'quorum' - query the majority of replicas, return values present in all of them - 'all' - query all replicas, and return values present in all replicas **kwargs: Any other named arguments to pass through to QdrantClient.search() Returns: List of Documents selected by maximal marginal relevance. rwrrrxrzr}r)r'max_marginal_relevance_search_by_vector r)rvrwrrrxrzr}rr+rs r%max_marginal_relevance_searchz$Qdrant.max_marginal_relevance_searchs[d++E22;t;   #'+#      r$c tK||d{V} |j| f|||||||d| d{VS)aReturn docs selected using the maximal marginal relevance. Maximal marginal relevance optimizes for similarity to query AND diversity among selected documents. Args: query: Text to look up documents similar to. k: Number of Documents to return. Defaults to 4. fetch_k: Number of Documents to fetch to pass to MMR algorithm. Defaults to 20. lambda_mult: Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5. filter: Filter by metadata. Defaults to None. search_params: Additional search params score_threshold: Define a minimal score threshold for the result. If defined, less similar results will not be returned. Score of the returned result might be higher or smaller than the threshold depending on the Distance function used. E.g. for cosine similarity only higher scores will be returned. consistency: Read consistency of the search. Defines how many replicas should be queried before returning the result. Values: - int - number of replicas to query, values should present in all queried replicas - 'majority' - query all replicas, but return values present in the majority of replicas - 'quorum' - query the majority of replicas, return values present in all of them - 'all' - query all replicas, and return values present in all replicas **kwargs: Any other named arguments to pass through to AsyncQdrantClient.Search(). Returns: List of Documents selected by maximal marginal relevance. Nr)r(amax_marginal_relevance_search_by_vectorrs r%amax_marginal_relevance_searchz%Qdrant.amax_marginal_relevance_search sh!% 2 25 9 9999999BTB   #'+#            r$c |j|f|||||||d| } tttd| S)aReturn docs selected using the maximal marginal relevance. Maximal marginal relevance optimizes for similarity to query AND diversity among selected documents. Args: embedding: Embedding to look up documents similar to. k: Number of Documents to return. Defaults to 4. fetch_k: Number of Documents to fetch to pass to MMR algorithm. lambda_mult: Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5. filter: Filter by metadata. Defaults to None. search_params: Additional search params score_threshold: Define a minimal score threshold for the result. If defined, less similar results will not be returned. Score of the returned result might be higher or smaller than the threshold depending on the Distance function used. E.g. for cosine similarity only higher scores will be returned. consistency: Read consistency of the search. Defines how many replicas should be queried before returning the result. Values: - int - number of replicas to query, values should present in all queried replicas - 'majority' - query all replicas, but return values present in the majority of replicas - 'quorum' - query the majority of replicas, return values present in all of them - 'all' - query all replicas, and return values present in all replicas **kwargs: Any other named arguments to pass through to QdrantClient.search() Returns: List of Documents selected by maximal marginal relevance. rr)2max_marginal_relevance_search_with_score_by_vectorrrr r)rrwrrrxrzr}rr+rs r%rz.Qdrant.max_marginal_relevance_search_by_vectorMsfbJ$I   #'+#      C 1 w//000r$c K|j|f|||||||d| d{V} tttd| S)a9Return docs selected using the maximal marginal relevance. Maximal marginal relevance optimizes for similarity to query AND diversity among selected documents. Args: query: Text to look up documents similar to. k: Number of Documents to return. Defaults to 4. fetch_k: Number of Documents to fetch to pass to MMR algorithm. Defaults to 20. lambda_mult: Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5. filter: Filter by metadata. Defaults to None. search_params: Additional search params score_threshold: Define a minimal score threshold for the result. If defined, less similar results will not be returned. Score of the returned result might be higher or smaller than the threshold depending on the Distance function used. E.g. for cosine similarity only higher scores will be returned. consistency: Read consistency of the search. Defines how many replicas should be queried before returning the result. Values: - int - number of replicas to query, values should present in all queried replicas - 'majority' - query all replicas, but return values present in the majority of replicas - 'quorum' - query the majority of replicas, return values present in all of them - 'all' - query all replicas, and return values present in all replicas **kwargs: Any other named arguments to pass through to AsyncQdrantClient.Search(). Returns: List of Documents selected by maximal marginal relevance and distance for each. rNr)3amax_marginal_relevance_search_with_score_by_vectorrrrrs r%rz/Qdrant.amax_marginal_relevance_search_by_vectorsfQP   #'+#            C 1 w//000r$c  |} j j| f} jjdj| |||dd||d |  fd D} t t j|| ||} fd| DS)a#Return docs selected using the maximal marginal relevance. Maximal marginal relevance optimizes for similarity to query AND diversity among selected documents. Args: query: Text to look up documents similar to. k: Number of Documents to return. Defaults to 4. fetch_k: Number of Documents to fetch to pass to MMR algorithm. Defaults to 20. lambda_mult: Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5. filter: Filter by metadata. Defaults to None. search_params: Additional search params score_threshold: Define a minimal score threshold for the result. If defined, less similar results will not be returned. Score of the returned result might be higher or smaller than the threshold depending on the Distance function used. E.g. for cosine similarity only higher scores will be returned. consistency: Read consistency of the search. Defines how many replicas should be queried before returning the result. Values: - int - number of replicas to query, values should present in all queried replicas - 'majority' - query all replicas, but return values present in the majority of replicas - 'quorum' - query the majority of replicas, return values present in all of them - 'all' - query all replicas, and return values present in all replicas **kwargs: Any other named arguments to pass through to QdrantClient.search() Returns: List of Documents selected by maximal marginal relevance and distance for each. NT rCrrrzrrrr}rcjg|]/}j|jjn|j0SrYrIvectorgetrs r%rzMQdrant.max_marginal_relevance_search_with_score_by_vector.. Q   + M  d. / / /   r$rwrcg|]<}|jjj|jf=Sr#rrirr)s r%rzMQdrant.max_marginal_relevance_search_with_score_by_vector..e    00AJ(,-         r$r#)rIrBrrCrnparrayr)rrwrrrxrzr}rr+rrD mmr_selectedrs` @r%rz9Qdrant.max_marginal_relevance_search_with_score_by_vectorsb!   ' ,l;L$$+$   0%'+#          "    2 HY  qk         "    r$c j Kjtjjtrt d|} j j| f} jjdj| |||dd||d | d{V fd D} ttj || ||} fd| DS) a/Return docs selected using the maximal marginal relevance. Maximal marginal relevance optimizes for similarity to query AND diversity among selected documents. Args: query: Text to look up documents similar to. k: Number of Documents to return. Defaults to 4. fetch_k: Number of Documents to fetch to pass to MMR algorithm. Defaults to 20. lambda_mult: Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5. Returns: List of Documents selected by maximal marginal relevance and distance for each. NrqTrcjg|]/}j|jjn|j0SrYrrs r%rzNQdrant.amax_marginal_relevance_search_with_score_by_vector..Urr$rcg|]<}|jjj|jf=Sr#rrs r%rzNQdrant.amax_marginal_relevance_search_with_score_by_vector..^rr$r#) rJrOrrrr.rIrrCrrrrs` @r%rz:Qdrant.amax_marginal_relevance_search_with_score_by_vector"sN:   $   %'7) ) $&M !   ' ,l;L0)0   0%'+#                "    2 HY  qk         "    r$Optional[List[str]]Optional[bool]c x|j|j|}|jtjjkS)Delete by vector ID or other criteria. Args: ids: List of ids to delete. **kwargs: Other keyword arguments that subclasses might use. Returns: True if deletion is successful, False otherwise. rCpoints_selector)rBdeleterCstatusr UpdateStatus COMPLETEDr)r`r+rs r%rz Qdrant.deleteks>## 0$  } 3 ===r$cK|jt|jjtrt d|j|j|d{V}|jtj j kS)rNrqr) rJrOrrrr.rrCrrrrrs r%adeletezQdrant.adelete|s   $   %'7) ) $&M (// 00         } 3 ===r$FCosinecls Type[Qdrant]rlocationurlport Optional[int] grpc_port prefer_grpcboolhttpsapi_keyprefixtimeouthostpath distance_func shard_numberreplication_factorwrite_consistency_factoron_disk_payload hnsw_configOptional[models.HnswConfigDiff]optimizers_config%Optional[models.OptimizersConfigDiff] wal_configOptional[models.WalConfigDiff]quantization_config#Optional[models.QuantizationConfig] init_fromOptional[models.InitFrom]on_diskforce_recreatec! |j||||||| | | | | |||||||||||||||||| fi|!}"|"|||||"S)aConstruct Qdrant wrapper from a list of texts. Args: texts: A list of texts to be indexed in Qdrant. embedding: A subclass of `Embeddings`, responsible for text vectorization. metadatas: An optional list of metadata. If provided it has to be of the same length as a list of texts. ids: Optional list of ids to associate with the texts. Ids have to be uuid-like strings. location: If ':memory:' - use in-memory Qdrant instance. If `str` - use it as a `url` parameter. If `None` - fallback to relying on `host` and `port` parameters. url: either host or str of "Optional[scheme], host, Optional[port], Optional[prefix]". Default: `None` port: Port of the REST API interface. Default: 6333 grpc_port: Port of the gRPC interface. Default: 6334 prefer_grpc: If true - use gPRC interface whenever possible in custom methods. Default: False https: If true - use HTTPS(SSL) protocol. Default: None api_key: API key for authentication in Qdrant Cloud. Default: None Can also be set via environment variable `QDRANT_API_KEY`. prefix: If not None - add prefix to the REST URL path. Example: service/v1 will result in http://localhost:6333/service/v1/{qdrant-endpoint} for REST API. Default: None timeout: Timeout for REST and gRPC API requests. Default: 5.0 seconds for REST and unlimited for gRPC host: Host name of Qdrant service. If url and host are None, set to 'localhost'. Default: None path: Path in which the vectors will be stored while using local mode. Default: None collection_name: Name of the Qdrant collection to be used. If not provided, it will be created randomly. Default: None distance_func: Distance function. One of: "Cosine" / "Euclid" / "Dot". Default: "Cosine" content_payload_key: A payload key used to store the content of the document. Default: "page_content" metadata_payload_key: A payload key used to store the metadata of the document. Default: "metadata" vector_name: Name of the vector to be used internally in Qdrant. Default: None batch_size: How many vectors upload per-request. Default: 64 shard_number: Number of shards in collection. Default is 1, minimum is 1. replication_factor: Replication factor for collection. Default is 1, minimum is 1. Defines how many copies of each shard will be created. Have effect only in distributed mode. write_consistency_factor: Write consistency factor for collection. Default is 1, minimum is 1. Defines how many replicas should apply the operation for us to consider it successful. Increasing this number will make the collection more resilient to inconsistencies, but will also make it fail if not enough replicas are available. Does not have any performance impact. Have effect only in distributed mode. on_disk_payload: If true - point`s payload will not be stored in memory. It will be read from the disk every time it is requested. This setting saves RAM by (slightly) increasing the response time. Note: those payload values that are involved in filtering and are indexed - remain in RAM. hnsw_config: Params for HNSW index optimizers_config: Params for optimizer wal_config: Params for Write-Ahead-Log quantization_config: Params for quantization, if None - quantization will be disabled init_from: Use data stored in another collection to initialize this collection force_recreate: Force recreating the collection **kwargs: Additional arguments passed directly into REST client initialization This is a user-friendly interface that: 1. Creates embeddings, one for each text 2. Initializes the Qdrant database as an in-memory docstore by default (and overridable to a remote docstore) 3. Adds the text embeddings to the Qdrant database This is intended to be a quick way to get started. Example: .. code-block:: python from langchain_qdrant import Qdrant from langchain_openai import OpenAIEmbeddings embeddings = OpenAIEmbeddings() qdrant = Qdrant.from_texts(texts, embeddings, "localhost") )construct_instancero#rr\rr^r`rrrrrrrrrrrrCrrFrGrIrbrrrrrrrrrrrr+qdrants# r% from_textszQdrant.from_textss\('                    $        ;  <=  @  3 ;;; r$c |td|jd|||||| | | | | |d |\}}|||||||||S)z Get instance of an existing Qdrant collection. This method will return the instance of the store without inserting any new embeddings Nz,Must specify collection_name. Received None. rrrrrrrrrrr)rBrJrCrDrFrGrHrIr#)rP_generate_clients)rrrrCrrrrrrrrrrrFrGrHrIr+rBrJs r%from_existing_collectionzQdrant.from_existing_collectionHs8  "KLL L4s4 #   s%+ 3!5/#    r$c!K|j||||||| | | | | |||||||||||||||||| fi|!d{V}"|"||||d{V|"S)aConstruct Qdrant wrapper from a list of texts. Args: texts: A list of texts to be indexed in Qdrant. embedding: A subclass of `Embeddings`, responsible for text vectorization. metadatas: An optional list of metadata. If provided it has to be of the same length as a list of texts. ids: Optional list of ids to associate with the texts. Ids have to be uuid-like strings. location: If ':memory:' - use in-memory Qdrant instance. If `str` - use it as a `url` parameter. If `None` - fallback to relying on `host` and `port` parameters. url: either host or str of "Optional[scheme], host, Optional[port], Optional[prefix]". Default: `None` port: Port of the REST API interface. Default: 6333 grpc_port: Port of the gRPC interface. Default: 6334 prefer_grpc: If true - use gPRC interface whenever possible in custom methods. Default: False https: If true - use HTTPS(SSL) protocol. Default: None api_key: API key for authentication in Qdrant Cloud. Default: None Can also be set via environment variable `QDRANT_API_KEY`. prefix: If not None - add prefix to the REST URL path. Example: service/v1 will result in http://localhost:6333/service/v1/{qdrant-endpoint} for REST API. Default: None timeout: Timeout for REST and gRPC API requests. Default: 5.0 seconds for REST and unlimited for gRPC host: Host name of Qdrant service. If url and host are None, set to 'localhost'. Default: None path: Path in which the vectors will be stored while using local mode. Default: None collection_name: Name of the Qdrant collection to be used. If not provided, it will be created randomly. Default: None distance_func: Distance function. One of: "Cosine" / "Euclid" / "Dot". Default: "Cosine" content_payload_key: A payload key used to store the content of the document. Default: "page_content" metadata_payload_key: A payload key used to store the metadata of the document. Default: "metadata" vector_name: Name of the vector to be used internally in Qdrant. Default: None batch_size: How many vectors upload per-request. Default: 64 shard_number: Number of shards in collection. Default is 1, minimum is 1. replication_factor: Replication factor for collection. Default is 1, minimum is 1. Defines how many copies of each shard will be created. Have effect only in distributed mode. write_consistency_factor: Write consistency factor for collection. Default is 1, minimum is 1. Defines how many replicas should apply the operation for us to consider it successful. Increasing this number will make the collection more resilient to inconsistencies, but will also make it fail if not enough replicas are available. Does not have any performance impact. Have effect only in distributed mode. on_disk_payload: If true - point`s payload will not be stored in memory. It will be read from the disk every time it is requested. This setting saves RAM by (slightly) increasing the response time. Note: those payload values that are involved in filtering and are indexed - remain in RAM. hnsw_config: Params for HNSW index optimizers_config: Params for optimizer wal_config: Params for Write-Ahead-Log quantization_config: Params for quantization, if None - quantization will be disabled init_from: Use data stored in another collection to initialize this collection force_recreate: Force recreating the collection **kwargs: Additional arguments passed directly into REST client initialization This is a user-friendly interface that: 1. Creates embeddings, one for each text 2. Initializes the Qdrant database as an in-memory docstore by default (and overridable to a remote docstore) 3. Adds the text embeddings to the Qdrant database This is intended to be a quick way to get started. Example: .. code-block:: python from langchain_qdrant import Qdrant from langchain_openai import OpenAIEmbeddings embeddings = OpenAIEmbeddings() qdrant = await Qdrant.afrom_texts(texts, embeddings, "localhost") N)aconstruct_instancertr s# r% afrom_textszQdrant.afrom_textss^/s.                    $        ;  <=        @y#zBBBBBBBBB r$c z||dd}t|d} |ptjj}|}|jd||||||| | | | | d |\}!}"|!|}#|#r|r|!|d}#|#r|! |}$|$j j j }%t|%trZ|X||%vr>td|d|dd |%d |%|}%n|t|%tr=|;td|d d |%d t|%ts|td|d t|%t&jsJdt+|%|%j| krtd|%jd| d|%jj}&|&|krtd|&d|d|&dnNt'j| t&j||}'|||'i}'|!||'||||||||||  ||!|||||||"}(|(S)Nr-rrFrCExisting Qdrant collection  does not contain vector named ,. Did you mean one of the existing vectors: , S? If you want to recreate the collection, set `force_recreate` parameter to `True`.m uses named vectors. If you want to reuse it, please set `vector_name` to any of the existing named vectors: R.If you want to recreate the collection, set `force_recreate` parameter to `True`. doesn't use named vectors. If you want to reuse it, please set `vector_name` to `None`. If you want to recreate the collection, set `force_recreate` parameter to `True`.QExpected current_vector_config to be an instance of models.VectorParams, but got :Existing Qdrant collection is configured for vectors with % dimensions. Selected embeddings are _-dimensional. If you want to recreate the collection, set `force_recreate` parameter to `True`.-Existing Qdrant collection is configured for z similarity, but requested z+. Please set `distance_func` parameter to `l` if you want to reuse it. If you want to recreate the collection, set `force_recreate` parameter to `True`.sizedistancer rCvectors_configrrrrrrrrrrrBrCrDrFrGrHrIrJr#)embed_documentslenuuiduuid4hexrVrcollection_existsdelete_collectionget_collectionconfigparamsvectorsrOrrjoinkeysrr VectorParamsrQr&r'nameDistancecreate_collection)rr\rrrrrrrrrrrrrCrrFrGrIrrrrrrrrrrrr+partial_embeddings vector_sizerBrJr0collection_infocurrent_vector_configcurrent_distance_funcr)r s) r%r zQdrant.construct_instance2sF'66uRaRyAA,Q/00 )=TZ\\-=%++-- 4s4 #   #44_EE  & &  $ $_ 5 5 5 %  W %33O3TTO$3$:$A$I !/66 ;;R&;;;)0o000;00-1YY7L7Q7Q7S7S-T-T000)>(A(A+(N(N%%1488 [=P%,/,,yy!6!;!;!=!=>>,,,4d;; @K@W%=/=== 3V5HII  N045J0K0KNN    %)[88%,,1,,;F,,,&.399;; "% 55%,,,,$,,.,,,6$0 7N&"  $ $ /-)#5)A /'"3%$7# %   + 3!5+#%     r$cK||ddd{V}t|d} |ptjj}|}|jd||||||| | | | | d |\}!}"|!|}#|#r|r|!|d}#|#r|! |}$|$j j j }%t|%trZ|X||%vr>td|d|dd |%d |%|}%n|t|%tr=|;td|d d |%d t|%ts|td|d t|%t&jsJdt+|%|%j| krtd|%jd| d|%jj}&|&|krtd|%jd|dnNt'j| t&j||}'|||'i}'|!||'||||||||||  ||!|||||||"}(|(S)Nr-rrFrrrrrrrrrrr r!r"r#z6 similarity. Please set `distance_func` parameter to `r$r%r(r*r#)aembed_documentsr,r-r.r/rVrr0r1r2r3r4r5rOrrr6r7rrr8rQr&r'r9r:r;r<s) r%rzQdrant.aconstruct_instances!F$-#=#=eBQBi#H#HHHHHHH,Q/00 )=TZ\\-=%++-- 4s4 #   #44_EE  & &  $ $_ 5 5 5 %  Y %33O3TTO$3$:$A$I !/66 ;;R&;;;)0o000;00-1YY7L7Q7Q7S7S-T-T000)>(A(A+(N(N%%1488 [=P%,/,,yy!6!;!;!=!=>>,,,4d;; @K@W%=/===3V5HII  N045J0K0KNN    %)[88%,,1,,;F,,,&.399;; "% 55%,5&6$0 7N&"  $ $ /-)#5)A /'"3%$7# %   + 3!5+#%     r$r'c|dzdz S)z4Normalize the distance to a score on a scale [0, 1].g?g@r#)r's r%_cosine_relevance_score_fnz!Qdrant._cosine_relevance_score_fnss3#%%r$Callable[[float], float]c|jdkr|jS|jdkr|jS|jdkr|jSt d)a8 The 'correct' relevance function may differ depending on a few things, including: - the distance / similarity metric used by the VectorStore - the scale of your embeddings (OpenAI's are unit normed. Many others are not!) - embedding dimensionality - etc. rADOTEUCLIDzJUnknown distance strategy, must be cosine, max_inner_product, or euclidean)rHrE%_max_inner_product_relevance_score_fn_euclidean_relevance_score_fnrPrZs r%_select_relevance_score_fnz!Qdrant._select_relevance_score_fnxs`  !X - -2 2  #u , ,= =  #x / /5 52 r$c  |j||fi|S)Return docs and relevance scores in the range [0, 1]. 0 is dissimilar, 1 is most similar. Args: query: input text k: Number of Documents to return. Defaults to 4. **kwargs: kwargs to be passed to similarity search. Should include: score_threshold: Optional, a floating point value between 0 to 1 to filter the resulting set of retrieved docs Returns: List of Tuples of (doc, similarity_score) )rr)rvrwr+s r%(_similarity_search_with_relevance_scoresz/Qdrant._similarity_search_with_relevance_scoress!(1t0DDVDDDr$c0K|j||fi|d{VS)rNN)rrOs r%)_asimilarity_search_with_relevance_scoresz0Qdrant._asimilarity_search_with_relevance_scoress7*8T7qKKFKKKKKKKKKr$ List[dict]cg}t|D];\}}|td|||nd}|||||i<|S)NzpAt least one of the texts is None. Please remove it before calling .from_texts or .add_texts on Qdrant instance.) enumeraterPappend) rr\r^rFrGpayloadsrtextr=s r%_build_payloadszQdrant._build_payloadss ''  GAt| L(1'.sTC!%!6!6sE!B!Br$)must)rFilterrg)r)rxs` r%rzQdrant._qdrant_filter_from_dictsY 4}"(,,..    r$c|j|j|}n,|j||}ntdt |dr|n|S)zEmbed query text. Used to provide backward compatibility with `embedding_function` argument. Args: query: Query text. Returns: List of floats representing the query embedding. N2Neither of embeddings or embedding_function is settolist)rD embed_queryrSrPhasattrrwr)rvrs r%rzQdrant._embed_querysv ? &33E::II(4 55e<<  !UVVV%,Y%A%APy!!!yPr$cK|j!|j|d{V}n,|j||}ntdt |dr|n|S)zEmbed query text asynchronously. Used to provide backward compatibility with `embedding_function` argument. Args: query: Query text. Returns: List of floats representing the query embedding. Nrvrw)rD aembed_queryrSrPryrwrzs r%rzQdrant._aembed_querys ? &"o::5AAAAAAAAII(4 55e<<  !UVVV%,Y%A%APy!!!yPr$List[List[float]]c|jL|jt|}t|dr|}nl|jVg}|D]P}||}t|dr|}||Qntd|SzEmbed search texts. Used to provide backward compatibility with `embedding_function` argument. Args: texts: Iterable of texts to embed. Returns: List of floats representing the texts embedding. Nrwrv)rDr+rryrwrSrVrPr)r\rDrXrs r% _embed_textszQdrant._embed_texts0s ? &88eEEJz8,, 1'..00  & 2J - - 55d;; :x003 ) 0 0 2 2I!!),,,,  - QRR Rr$cK|jR|jt|d{V}t|dr|}nl|jVg}|D]P}||}t|dr|}||Qntd|Sr)rDrCrryrwrSrVrPrs r% _aembed_textszQdrant._aembed_textsKs ? &#??U LLLLLLLLJz8,, 1'..00  & 2J - - 55d;; :x003 ) 0 0 2 2I!!),,,,  - QRR Rr$AGenerator[Tuple[List[str], List[models.PointStruct]], None, None]c #DKt|}t|pg}t|pdt|D}tt||x}rtt||pd} tt||} |} fdt | | || jjD} | | fVtt||x}dSdS)Nc>g|]}tjjSr#r-r.r/r_s r%rz1Qdrant._generate_rest_batches..o!#J#J#JDJLL$4#J#J#Jr$cfg|]-\}}}tj|j|nj|i|.SN)r`rr_r PointStructrIrpoint_idrr_r)s r%rz1Qdrant._generate_rest_batches..xd.Hfg"'/"6*F3# r$)iterrrrziprYrFrG r)r\r^r`rbtexts_iteratormetadatas_iterator ids_iterator batch_textsbatch_metadatasrnbatch_embeddingsrhs ` r%rizQdrant._generate_rest_batchesfs]e!)/r22CJ#J#Jd5kk#J#J#JKK !&"D"DEEEk $"6*>I $00== 25$((#'01  2 2F(V# # # #9"&"D"DEEEk $ $ $ $ $r$@AsyncGenerator[Tuple[List[str], List[models.PointStruct]], None]c RKt|}t|pg}t|pdt|D}tt||x}rtt||pd} tt||} |d{V} fdt | | || jjD} | | fWVtt||x}dSdS)Nc>g|]}tjjSr#rrs r%rz2Qdrant._agenerate_rest_batches..rr$cfg|]-\}}}tj|j|nj|i|.Srrrs r%rz2Qdrant._agenerate_rest_batches..rr$)rrrrrrYrFrGrs ` r%rszQdrant._agenerate_rest_batchessre!)/r22CJ#J#Jd5kk#J#J#JKK !&"D"DEEEk $"6*>I&*%7%7 %D%DDDDDDD 25$((#'01  2 2F(V# # # # #9"&"D"DEEEk $ $ $ $ $r$0Tuple[QdrantClient, Optional[AsyncQdrantClient]]c  |tjd}td|||||||||| | d | } |dks| d} ntd|||||||||| | d | } | | fS)NQDRANT_API_KEYrz:memory:r#)osgetenvrr)rrrrrrrrrrrr+ sync_clientrJs r%rzQdrant._generate_clientss ?i 011G"  #       z ! !T%5 LL, !#'    LL((r$)rBrrCr;rDrErFr;rGr;rHr;rIr?rJrKrLrM)r'rE)NNr[) r\r]r^r_r`rarbrcr+rr'rd)ruNNrNN)rvr;rwrcrxryrzr{r|rcr}r~rrr+rr'r)ruN) rvr;rwrcrxryr+rr'r)rvr;rwrcrxryrzr{r|rcr}r~rrr+rr'r)rrrwrcrxryrzr{r|rcr}r~rrr+rr'r)rrrwrcrxryrzr{r|rcr}r~rrr+rr'r)rurrNNNN)rvr;rwrcrrcrrrxryrzr{r}r~rrr+rr'r)rrrwrcrrcrrrxryrzr{r}r~rrr+rr'r)rrrwrcrrcrrrxryrzr{r}r~rrr+rr'rrY)r`rr+rr'r)Frrr\rdrrr^r_r`rarr?rr?rrrrcrrrrrr?rr?rrrr?rr?rCr?rr;rFr;rGr;rIr?rbrcrrrrrrrrrrrrrrrrrrrrrrr+rr'r9)(rrrrrr?rCr?rr?rr?rrrrcrrrrrr?rr?rrrr?rFr;rGr;rHr;rIr?r+rr'r9)@rrr\rdrrrr?rr?rrrrcrrrrrr?rr?rrrr?rr?rCr?rr;rFr;rGr;rIr?rrrrrrrrrrrrrrrrrrrrrrr+rr'r9)r'rr'r)r'rF)ru)rvr;rwrcr+rr'r) r\r]r^r_rFr;rGr;r'rS) rZrrCr;rFr;rGr;r'r)rar;rbrr'rc)rxrnr'ro)rvr;r'r)r\r]r'r}) r\r]r^r_r`rarbrcr'r) r\r]r^r_r`rarbrcr'r) NNrrFNNNNNN)rr?rr?rrrrcrrrrrr?rr?rrrr?rr?r+rr'r)4rr r!r"r<__annotations__r>r@rWpropertyrDror3rtrrrrrrrrrrrrrrrr classmethodr rrr r staticmethodrErLrPrRrYrrhrrrrrrirsrr#r$r%r9r9Es}  &K%%%%"L""""!%K%%%% ,0#.$0!)%0&*15;;;;;;;;;;z   X +/'+     D+/'+ '''''X+/7;+/8<7171717171r+/ 11111*+/7;+/8<6 6 6 6 6 p+/7;+/8<8 8 8 8 8 z+/7;+/8<7171717171r+/7;+/8<8181818181z+/7;+/8<T T T T T l+/7;+/8<\ \ \ \ \ B +/7;+/8<= = = = = ~ +/7;+/8<> > > > > F +/7;+/8<<1<1<1<1<1| +/7;+/8<=1=1=1=1=1D +/7;+/8<U U U U U n +/7;+/8<F F F F F P>>>>>")->>>>>4 +/'+"&!"! $!% $!%"")-%#.$0%0&*,026*.7;CG59CG/3"&$Cnnnn[n`#)-"&!"! $!% $!%"#.$0!)%0%5 5 5 5 [5 n +/'+"&!"! $!% $!%"")-%#.$0%0&*,026*.7;CG59CG/3"&$Cnnnn[n` #'!"! $!% $!%"")-%#.$0%0&*,026*.7;CG59CG/3"&$=\\\\[\| #'!"! $!% $!%"")-%#.$0%0&*,026*.7;CG59CG/3"&$=____[_B&&&\&2EEEEE,LLLLL,[0    [  ,     QQQQ(QQQQ(6<+/'+ &$&$&$&$&$V+/'+ &$&$&$&$&$P"&!"! $!% $!%""4)4)4)4)\4)4)4)r$r9)r&r r'r )9 __future__rr1rr-rT itertoolsroperatorrtypingrrrr r r r r rrrrrnumpyrlangchain_core._api.deprecationrlangchain_core.documentsrlangchain_core.embeddingsrlangchain_core.runnables.configrlangchain_core.vectorstoresr qdrant_clientrrqdrant_client.httpr&qdrant_client.local.async_qdrant_localrlangchain_qdrant._utilsrr;rcrrr DictFilterrtMetadataFilter Exceptionrr3r9r#r$r%rs""""""  666666------000000;;;;;;33333399999999%%%%%%CCCCCC>>>>>>6c5c4t!;<<=J:v}45N'''''i'''. '':GLLLe")e")e")e")e")[e")e")MLe")e")e")r$