I'm a professional working in a heavy text-based industry. I'm looking to build a small homelab to tinker around with a local LLM specifically designed with my working data to see if it helps me with my everyday work. I can't use cloud services because of the sensitivity of my data.
I'm a tech-focused person and can program a little, also somewhat familiar with Python and R. But I don't have the spare time next to my job to set up the homelab all by myself.
These are the steps I need your help for:
1) Advice on buying the hardware for the local LLM.
2) Discussing a suited solution to use a good pre-trained model and train it further with my own data.
3) Deploy the solution to the hardware of step 1 (remotely, of course).
4) Creating a small instruction set on how to use the solution (further train the model, use my existing data, etc.).
5) Set up a front-end for easy interaction with the local LLM.
6) Be ready to answer questions for a couple of weeks.
It can (and probably should) all be open source solutions. That's it for now. If it works out well there might be very likely more work to come.
Location: Anywhere
Posted: Sept. 9, 2024, 9:44 a.m.
Apply Now Company Website