Anyone in tech confirm?

You're viewing a single thread.

165 comments

Not really. But the more you understand tech, the less you like the high tech market solutions and have the urge to have the same but in opensource.
I kinda dream that one day setups for mid-power home AI racks would be affordable enough to drop all the proprietary AIs and start automating my daily routines with agents & helping me build my local knowledge vector database with AI embeddings. No way in hell I'm exposing anything like that to thirdparties.
- The efficiency improvements in some open models are becoming crazy, like hundreds of times from a year ago. I have a setup such as yours on my framework which can handle a 120b param model fully loaded. It's capable of the RAG setup you are already envisioning.
  
  Moar deets, please!
  
  What do you mean framework? What are the specs?
  
  Hi,
  I'm running the AMD AI 395+ with the 128gb of unified RAM option.
  https://frame.work/ca/en/products/desktop-diy-amd-aimax300/configuration/new
  I love mine. I'm running NixOS.
  
  If you are capable of running 120b, the mistral cli bigger open source model in agentic mode could be available to you.
  How much did your setup cost, if you don't mind to disclose such information?
  
  I mentioned my spec above, and I added a 2tb m2, noctua upgrade and a couple expansion modules for the front. I preordered it around 6 months back so before any of the craziness. It was around $3600 CAD
  Edit: also thank you. I have been busy making workflows and haven't looked at the models in a bit. I'll check out mistral
  
  That's a good value! I was worrying something running mid power LLMs would cost 10k+ today. Sure ram screwup nuked the field, but it's also not on the top list of my priorities. So it gives me hope that by the time I would be ready it would be in even better shape.

165 comments