hi i am doing an insert in the pinecone adding a metadata.i am replicating the metadata to each document i have obtained from a split using lang chain. There is a strange behavior with pinecone or langchain when i am doing upsert . 1 - i am not explicity adding the attribute text in the metadata and the insert is adding the text attribute to my metadata.
2 -When i query for similarity the metadata returned has the column text and the value of the column seems to be replicated for each document and the thresholds are different.
any idea what is happen?
here is my code
#Create List of Dictionary for each chunk for this page
metalist =
metadict= {}
#Create metadatalist
for t in range(len(texts)):
metadict["id"] = str(offerid)
metadict["tenantid"] = str(tenantid)
metadict["villageid"] = str(villageid)
metadict["sourcetype"] = "document"
metadict["type"] = str(eventtype)
metadict["actionid"] = str(actionid)
metadict["userfirstname"] = str(userfirstname)
metadict["title"] = str(title)
metadict["body"] = str(body)
metadict["tag"] = str(tag)
metadict["documentname"] = document_name
metalist.append(metadict)
docsearch = Pinecone.from_texts([t.page_content for t in texts], embeddings, index_name=PINE_INDEX_NAME,metadatas=metalist)