HOME

Hermes , omo , claud... codex... all have 1 thing in common.

They're agentic.

Well more than one thing but I was going for drama.
The truth is they have a lot in common but that's not what I'm discussing today!

Today I'm talking about HOME. My personal little project to house an AI , not a user interface but a user dashboard and at a glance system. Nodes determine whats being used and whats not. A postgres db houses it's memories, and it has a full UI like an entity map.

It's coming along amazingly - but man is it a pain in the ass this llm context hacks.
The layered file approach with calls depending on context. GREAT I cut the context window in half!
But I also have a 128k context window.. so I wonder, why is it so fucking slow to respond.

It's the cloud model, I havve 2.1 gbps download and 2.0 gbps upload.

So why is it so slow?
I load up a local model on my pitiful 4060 and it runs three times as fast as the cloud model.

I can't figure out where the bloat is coming from..
This is driving me nuts!!!

BUT!! Home is awesome, and looking great! Screenshots in the Community with a quick write up on the app. <3

Be your best
All the best

• h4 (b'.')b