I am now running my own local LLM server at home via Ollama. Playing with models but liking llama2 with all the AI safety features turned off. It's connected to my Obsidian knowledge base but want to augment (RAG) it a lot more. One custom gpt so far around Product Desigb. Can access via mobile when out of the home.

Memetic-Artist. Techno-Optimist. Purple. Energy. Nouns. Optimism Citizens House. ↑↑↑ Product @CharmVerse. 

Will hook up Stable diffusion to it soon and will write a few bespoke apps.

american computer programmer • cofounder @glif • https://jamiedubs.com • fid 336

You can just host it on a public webserver but I use NordVPN and they have a VPN Mesh feature where all your machines are on the same virtual network no matter where you are.

how are you accessing it remotely? Love idea of wiring it up to my Obsidian; especially if I can use on the go

🟣 Artist | VR • Creator of Purple Valley, 50M shares on AR • Graffiti Writer since 2010
https://linktr.ee/ottisots

Rio de Janeiro, State of Rio de Janeiro, Brazil

"...we knew this day would come, it was only a matter of time."

At first just because I could. Then when I did I was blown away by how much faster it is. Then played around and realized there is a lot more you can do with local documents, so feel like I can train it perfectly with time. Bonus its all open source, even the  models.

What you plan on doing with it? What made you do a homebrew?

Principal at Cobalt Ventures. Write @ TheItinerant.io ; BenErsing.com Host: /venturecapital /finance /adventures

I'm bullish on whatever Apple decides to do. I kind of thought you couldn't compete with the cloud and that's just not true. You can totally run them locally and they are even faster that way.

What's been the biggest learning in setting it up?

this is pretty awesome. is there a primer on how to set this up? and what kind of hardware do you need to do it?