Update app.py

app.py

@@ -11,7 +11,7 @@ vector_store = FAISS.load_local("db.index", embeddings=None, allow_dangerous_des
 client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 
 # Create an LLM object from the InferenceClient
-llm = HuggingFaceInferenceClient(client)
+llm = HuggingFaceTextGenInference(client)
 
 # Initialize the RetrievalQA chain
 qa = RetrievalQA.from_chain_type(llm=llm, chain_type="stuff", retriever=vector_store.as_retriever())