How to get response based on image?

[QiuyiF2089]

Hi guys, I try to use the
"
import openparse
from llama_index.core import VectorStoreIndex

doc_path = "./attention_is_all_you_need.pdf"
parser = openparse.DocumentParser()
parsed_doc = parser.parse(doc_path)

nodes = parsed_doc.to_llama_index_nodes()
index = VectorStoreIndex(nodes=nodes)

query_engine = index.as_query_engine()
response = query_engine.query("can you tell me the attention graph?")
print(response).
"
Does it work if I ask question like "can you tell me the attention graph?". I'm not sure whether the generated response is based on the graph or paragraphs.