Update utils.py

utils.py

@@ -425,7 +425,7 @@ def rag_chain(llm, prompt,  retriever):
         #RAG_CHAIN_PROMPT = PromptTemplate(template="Context: {context}\n\nQuestion: {question}\n\nAnswer:")
 
         # Inahlte Abrufen der relevanten Dokumente
-        doc_contents = [doc["page_content"] for doc in extracted_docs]
+        doc_contents = [doc["content"] for doc in extracted_docs]
 
         #Berechne die Ähnlichkeiten und finde das relevanteste Dokument
         question_embedding = embedder_modell.encode(prompt, convert_to_tensor=True)
@@ -437,7 +437,7 @@ def rag_chain(llm, prompt,  retriever):
         most_relevant_docs = [extracted_docs[i] for i in most_relevant_doc_indices]
 
         #Kombiniere die Inhalte aller relevanten Dokumente
-        combined_content = " ".join([doc["page_content"] for doc in most_relevant_docs])
+        combined_content = " ".join([doc["content"] for doc in most_relevant_docs])
 
         #Formuliere die Eingabe für das Generierungsmodell
         input_text = f"frage: {prompt} kontext: {combined_content}"