better pre-processing pdf files

[cloudrage999]

i think private GPT needs to better parse and pre-process the pdf files , maybe using unstructured io or OCR tools
now that we have lamaparse , i think we should add such things to privategpt
this will make a big difference in getting accurate answers
currently if you want to run privategpt locally, you wont have a good experience if you got bunch of complex pdf files that have tables,messy format,pictures,diagrams etc