How to manage repeated file conversions

What I’m trying to do

I am interested in experimenting with Chat with RTX on my vault. I plan to make a copy of all notes, then use PowerRename utility to change the extension from .md to .txt.

Is there a more reasonable way to utilize Chat with RTX without having to regularly make duplicates and convert file formats? I imagine a script could handle this, but I am having difficulty seeing that as a final solution.

Since I am going to be wanting to modify things, the text files will have to be reindexed by the LLM to be useful, regardless of their format. However, if the indexing begins to happen automatically, it sure would be nice to make modifications to my vault’s md files and have txt versions of those files replace the existing ones in the Chat with RTX directory.

Things I have tried

I searched “convert to .txt” and hoped to find a similar use case where the conversions were needed regularly on the same set of changing .md files, but had no luck.

I am also looking for a solution by searching for similar use cases with existing local LLMs.

Thanks in advance!

I don’t have a solution, but I do share the problem. I installed Chat with RTX last night and the first thing I did was to point it at my notes, and of course all it found was a PDF! I’m very surprised it supports .txt but not .md. Connecting a local LLM to notes seems like a great idea to me. My hope is that Nvidia will add .md support soon.

1 Like

Also very interested in this! I posted on nvidia’s dev forum regarding support for .md. Not a lot of attention yet…

https://forums.developer.nvidia.com/c/ai-data-science/ai-foundation-models/669 if you want to try and make it more visible with a +1 :slight_smile:

1 Like