+#hdZddlmZmZddlZddlZddlmZddl m Z  ddl Z e e dGddejZy#e $rdZ Y)wxYw) zxRetriever model for faiss: https://github.com/facebookresearch/faiss. Author: Jagane Sundar: https://github.com/jagane. )OptionalUnionN)SentenceTransformersVectorizer)dotdictzb The faiss package is required. Install it using `pip install dspy-ai[faiss-cpu]` creZdZdZddeffd Zd dZd deee efde ede jfdZ xZS) FaissRMaA retrieval module that uses an in-memory Faiss to return the top passages for a given query. Args: document_chunks: the input text chunks vectorizer: an object that is a subclass of BaseSentenceVectorizer k (int, optional): The number of top passages to retrieve. Defaults to 3. Returns: dspy.Prediction: An object containing the retrieved passages. Examples: Below is a code snippet that shows how to use this as the default retriver: ```python import dspy from dspy.retrieve import faiss_rm document_chunks = [ "The superbowl this year was played between the San Francisco 49ers and the Kanasas City Chiefs", "Pop corn is often served in a bowl", "The Rice Bowl is a Chinese Restaurant located in the city of Tucson, Arizona", "Mars is the fourth planet in the Solar System", "An aquarium is a place where children can learn about marine life", "The capital of the United States is Washington, D.C", "Rock and Roll musicians are honored by being inducted in the Rock and Roll Hall of Fame", "Music albums were published on Long Play Records in the 70s and 80s", "Sichuan cuisine is a spicy cuisine from central China", "The interest rates for mortgages is considered to be very high in 2024", ] frm = faiss_rm.FaissRM(document_chunks) turbo = dspy.OpenAI(model="gpt-3.5-turbo") dspy.settings.configure(lm=turbo, rm=frm) print(frm(["I am in the mood for Chinese food"])) ``` Below is a code snippet that shows how to use this in the forward() function of a module ```python self.retrieve = FaissRM(k=num_passages) ``` kc|r||_nt|_|j|}tj|}t |d}t j jd|t |dkr6tj||_ |jj|nid}tj|}tj||||_ |jj||jj|t j j|jjd||_t |E|y)zInits the faiss retriever. Args: document_chunks: a list of input strings. vectorizer: an object that is a subclass of BaseTransformersVectorizer. k: number of matches to return. rzFaissRM: embedding size=dz vectors in faiss index)r N) _vectorizerrnparraylendspyloggerinfofaiss IndexFlatL2 _faiss_indexadd IndexIVFFlattrainntotal_document_chunkssuper__init__) selfdocument_chunks vectorizerr embeddingsxbdnlist quantizer __class__s S/var/www/html/sandstorm/venv/lib/python3.12/site-packages/dspy/retrieve/faiss_rm.pyrzFaissRM.__init__Ds' )D =?D %%o6 XXj ! 1J 3A378 r7S= % 1 1! 4D     ! !" %E))!,I % 2 29a GD     # #B '    ! !" % D--4455LMN / 1returncHtt|D]}||}||}tjj d||tt|D]C}tjj d|d||d||d|j ||Ey)NzQuery: z Hit z = /z: )rangerrrdebugr)rqueries index_list distance_listiindices distancesjs r&_dump_raw_resultszFaissRM._dump_raw_resultsdss7|$ uA mG%a(I KK   |4 53w<( u !!HQCs71:,a ! ~RPTPePefmnofpPqOr"st u  u r'query_or_queriesc t|tr|gn|Dcgc]}|s| c}|j}tj|}t dk(rp|j j||xs |j\}}|dD cgc]} |j| | f} } | D cgc]} t| d| ddc} S|j j||xs |jdzfi|\}}i} tt |D]V} || }|| }t|xs |jdzD]+}||}||}|| vr| |j|&|g| |<-Xt| jfdd|xs |j}|Dcgc] \}}t|j||d"c}}Scc}wcc} wcc} wcc}}w)a Search the faiss index for k or self.k top passages for query. Args: query_or_queries (Union[str, List[str]]): The query or queries to search for. Returns: dspy.Prediction: An object containing the retrieved passages. r) long_textindexcTtt|dz t|dfS)Nr7)rsum)xr-s r&z!FaissRM.forward..s/G WZ[\]^[_W`H`befghifjbkGlr')keyN) isinstancestrr r rrrsearchr rrr+appendsorteditems)rr5r kwargsqr emb_npar/r.indpassagespassagepassage_scoresembr1r2resneighbordistancesorted_passages passage_index_r-s @r&forwardzFaissRM.forwardms)33CS(I#$O_%+1+%%g. ((:& w<1 (,(9(9(@(@!+tvv(V %M:EOPQ]Sc..s3S9SHS[cdPWG'!*wqzJKd d$~-"8,33H=08zN8,  : : !!5!5!7=lm ak466  %4  q $"7"7 "FQ^_ `  A,Td0 sG G G0G#%G)Nr:)r(N)N)__name__ __module__ __qualname____doc__intrr4rrAlistrr PredictionrT __classcell__)r%s@r&rrsM'RC@- c49n(=- (3-- eietet- r'r)rXtypingrrnumpyr rdsp.modules.sentence_vectorizerr dsp.utilsrr ImportErrorRetrieverr'r&rdsd# J =   @ dmm@  EsAA A