@contributor

try response.source_nodes[0].text ?

still the gives a blank output

Btw this might be the case why, the document i'm using is a scanned pdf

is there a way to work around it ?

oh that definitely could be why. Try using unstructured to load the document, or even better, llama-parse