Has anyone combined the jfk files into one text file for prompting?

Reply to this note

Please Login to reply.

Discussion

Good morning! 🌻💜

Are you sure you want to know?

Wow this was fast!

but can’t login with Ostrich 🤷‍♂️

I saw someone sharing a link few hours ago, but I don't remember who.

Nothing to see in those files.

FYI the third batch of the files just dropped a few hours ago , and they already started taking down some random pages from the previous batches

Can do easily enough.

I’ll make it available this evening if you like nostr:npub1r0rs5q2gk0e3dk3nlc7gnu378ec6cnlenqp8a3cjhyzu6f8k5sgs4sq9ac

That would be cool

What are you using to query it? Have been trying Open Web AI and Deepseek, but it seems to perform a search / query rather that have overall understanding of the content (even though it builds a vector DB).

Not sure yet

Here are the text files to put into an AI model:

https://github.com/BenGWeeks/JFKAssassinationRecords

Just download them and use goose + your LLM of choice

I think the need to be OCR'd first. And that's a bit challenging given all the stamps and scribbles in there. Someone needs to train an AI to OCR only the underlying doc.