Testing pdf image scraping

Prompting
,
1

testing pdf image scraping , found other alternative to resolve this. closing this thread,testing pdf image scraping , found other alternative to resolve this. closing this thread thanks thanks

1 Like
2

Hi,
I think you would need to take use of function call’s (https://platform.openai.com/docs/guides/function-calling) to set it up so that the LLM can return specific details, and also you can use something like RAG if it is a bigger document or just passing the document would be fine as well in the prompt it self. The responding part with the image’s and such would need some thinking but I don’t think it would be that hard to implement, you would need to specify which components need so the prompt would contain the attachments as well something like


The following are the content of the document (take it into context and produce an output):

This engine has xxx, it has a horse power of xxx

[Fig 1. Contains a Analysis of the Engine Performance]

You can use it and make it so the function call can select those figures, or related media, and then return it.

This is just a simple approach you can go for :slight_smile:

Hope this helps

1 Like