-
Notifications
You must be signed in to change notification settings - Fork 174
Open
Labels
P3bugSomething isn't workingSomething isn't workinginformation-neededInformation needed from the userInformation needed from the userintegration:pgvector
Description
I try to run the example in this tutorial:
https://docs.haystack.deepset.ai/docs/pgvectordocumentstore
But I cannot write to the document store:
KeyError Traceback (most recent call last)
Cell In[4], line 1
----> 1 document_store.write_documents([
2 Document(content="This is first", embedding=[0.1]*768),
3 Document(content="This is second", embedding=[0.3]*768)
4 ])
File ~/.../haystack_integrations/document_stores/pgvector/document_store.py:503, in PgvectorDocumentStore.write_documents(self, documents, policy)
500 if policy == DuplicatePolicy.NONE:
501 policy = DuplicatePolicy.FAIL
--> 503 db_documents = self._from_haystack_to_pg_documents(documents)
505 sql_insert = SQL(INSERT_STATEMENT).format(
506 schema_name=Identifier(self.schema_name), table_name=Identifier(self.table_name)
507 )
509 if policy == DuplicatePolicy.OVERWRITE:
File ~/.../haystack_integrations/document_stores/pgvector/document_store.py:559, in PgvectorDocumentStore._from_haystack_to_pg_documents(documents)
556 db_document["blob_meta"] = Jsonb(blob.meta) if blob and blob.meta else None
557 db_document["blob_mime_type"] = blob.mime_type if blob and blob.mime_type else None
--> 559 db_document["dataframe"] = Jsonb(db_document["dataframe"]) if db_document["dataframe"] else None
560 db_document["meta"] = Jsonb(db_document["meta"])
562 if "sparse_embedding" in db_document:
KeyError: 'dataframe'
haystack-ai==2.14.3
haystack-experimental==0.10.0
pgvector==0.4.1
pgvector-haystack==1.3.0
Metadata
Metadata
Assignees
Labels
P3bugSomething isn't workingSomething isn't workinginformation-neededInformation needed from the userInformation needed from the userintegration:pgvector