Context

· 1 min read

Context - Context, context, context...

Everything is context, as much in life as it is in the world of AI

So I'm trying to create a local db fueled AI.
I have the model based on Gemma4:e2b Cause we're working on a 8GB Vram system.

I have the code, system prompt compressed and truncated properly, I have a custom ingestion protocol, that has the model take bits of the db at a time and post them one after another to prevent the loss of context.

The issue?
I dunno who controls the context for a cloud model.
Is it the home of the model that does the heavy lifting and controls the context?
Or is the context window handled locally?

I dunno - I can't find anything on that in the documents.
I can't even find it online.

All I know is I keep getting this dumb error - where it craps out .

I also can't get the damn thing to stop narrating and answer organically lol

It's crazy -
Driving me a wee bit nuts.

But I'm good. Got the gumption today to do it to it.

It's been a rough few weeks. Very rough, with loss, failure, and bad news.
But this is giving me a little hope, this is me working towards my ultimate goal.
Project SYDK.

But that's a long ways away.

In any event ,
Be your best ,
All the best ,
- h4

More from the blog