its_me_xiphos

Hey all,

I'm on the last few months of my contract, in higher ed. While I try to find work that's more permanent, I do count myself lucky to be in a better position to offer help with little worry of reprucussions. In brief, I am trying to do whatever I can to support my LGBTQ+ students who are terrified. I'm a tough person, like I've seen some shit (war) but its breaking my soul feeling powerless here in the States.

Any suggestions, organizations, or other ideas I could have at the ready if a student needs help? I just want to be the best ally and empathetic human I can from my position, while I can.

I'm also being cautious since its getting harder to know who and what to trust, but ideas on exercising greater caution when offering help are welcome too. Sad I have to even convey that...

Been having too much fun using LLMs hosted locally, but can't seem to get Ollama's chat with documents to work well. Lots of "what are you talking about? There are no documents here" issues. Does anyone have any recommendations to either a) figure out what's going wrong or b) Alternative locally hosted options that chat with documents works well with (GPT4all or something?)

I was very excited to learn about this project...only to discover it's neither free nor open source. Does anyone know of any true open source and accessible tools for Syllabus sharing/curating/researching?

I'm new to the field of large language models (LLMs) and I'm really interested in learning how to train and use my own models for qualitative analysis. However, I'm not sure where to start or what resources would be most helpful for a complete beginner. Could anyone provide some guidance and advice on the best way to get started with LLM training and usage? Specifically, I'd appreciate insights on learning resources or tutorials, tips on preparing datasets, common pitfalls or challenges, and any other general advice or words of wisdom for someone just embarking on this journey.

Thanks!

Forgive my ignorance, but I've got a question concerning OCR tools. Until now, I have utilized a paid service to upload, scan, convert them to searchable documents, and store my handwritten Uni notes. Handwritten because, frankly, my brain seems to engage with the content "better" than by digital note-taking.

It worked fine for what I needed, so I have never investigated open-source or had actual ownership/control over my uploaded notes before. As my work expands and the database of notes grows, maintaining data privacy is a huge concern, and I do not want to use the same system for interviews and such. My Uni has been, well, unhelpful sadly.

Are there any recommendations for having a similar system that puts more control and privacy in my hands?