Spyke
Ollama·Ollama - Local LLMs for everyone!byAuster

How to provide local files for Ollama models to use on replies?

How to provide local files for Ollama models to use on replies?

For individual files, apparently #Ollama allows natively from what I could find, but I couldn't understand the instructions. Is there some tutorial I may have missed that explains it in layman's terms?

I also remember someone on the threadiverse posting some tool or plugin for that too, which I'll try looking for again when I'm not busy.

And what about whole directories, would it be possible without converting everything into a single file if the folder has only/mostly TXT files or similar?

Thanks in advance, and aldo thanks for the patience with the newbie questions! "<.<

@[email protected]

View original on thebrainbin.org
slrpnk.net

You might like Open Web UI or something similar. They have a RAG (retrieval augmented generation) feature where you can upload and embed a whole folder of info and the LLM will search that for additional context before answering your question. Connects right up to ollama.

Edit: or AnythingLLM is another really popular one that I have tried and worked well for me

2

You reached the end

How to provide local files for Ollama models to use on replies? | Spyke