Extract Table Info From SCANNED PDF & Summarise It Using Llama3.1 via Ollama | LangChain

[@Jeganbaskaran]

Thank you for this video. It would be good if you try the same with image as well. Images are not extracted properly on scanned copy. can you recommend any other packages help to extract images even better?

[@datasciencebasics]

You are welcome. Did you try Llamaparse from LlamaIndex ? It might be good.
Super Easy Way To Parse PDF | LlamaParse From LlamaIndex | LlamaCloud
th-cam.com/video/wRMnHbiz5ck/w-d-xo.html

[@stanTrX]

Thanks. I think unstructured is not open source. Can you suggest any pdf to data library which is completely free, such as tabula-py or pdfplumber? Have you tested with these or anything else which performs better?

[@datasciencebasics]

Unstructured has paid version via api but you can use pip to install but might not have all functionalities. Nope haven’t tried other packages.

[@arpittalmale6440]

Sir can you make a video on LangGraph and for Agents...

[@datasciencebasics]

I have created 2 videos about Langgraph, you can start feom this one,
Introduction to LangGraph: Building and Enhancing LLM Agents
th-cam.com/video/eOhTv5WhvSM/w-d-xo.html

[@arpittalmale6440]

@@datasciencebasics ok thank you sir

[@riyazvali5798]

This is possible to try on colab

[@datasciencebasics]

yes, it should be. might be difficult if colab doesn’t allow to install some system packages but give a try.