my llm object is like this -> llm =

from llama_index import ServiceContext
service_context = ServiceContext.from_defaults(llm=llm, embed_model=embed_model)

get_response_synthesizer(...., service_context=service_context)