in the evaluation script the final output is document name instead we need context/content from all the similar document
in the evaluation script the final output is document name instead we need context/content from all the similar document