fix condition to establish when we need to cleanup the storage

lfoppiano · Nov 29, 2023 · ba399dc · ba399dc
1 parent 153c004
commit ba399dc
Showing 1 changed file with 2 additions and 1 deletion.
diff --git a/document_qa/document_qa_engine.py b/document_qa/document_qa_engine.py
@@ -251,7 +251,8 @@ def create_memory_embeddings(self, pdf_path, doc_id=None, chunk_size=500, perc_o
         else:
             hash = metadata[0]['hash']
 
-        if hash not in self.embeddings_dict.keys():
+        if hash not in self.embeddings_dict.keys() and (
+                'documents' in self.embeddings_dict.get() and len(self.embeddings_dict.get()['documents']) == 0):
             self.embeddings_dict[hash] = Chroma.from_texts(texts,
                                                            embedding=self.embedding_function,
                                                            metadatas=metadata,