prefix caching in llama.cpp: worth it?

[marbletrack]

A couple questions before I guess too hard. What model, what runtime, and roughly what context size?

[dockridge]

A couple questions before I guess too hard. What model, what runtime, and roughly what context size?

That matches what I was seeing. I don't think your conclusion is far off.

[rocmriley]

What worked for me was pretty simple. Before rebuilding anything, I'd test one clean baseline with Llama and compare from there. Not glamorous advice, but repeatable baselines save a lot of time.

[embershed99]

A couple things I'd check first. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at small local stacks. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.

[vrambandit]

Ran a quick check on my box. My result was mixed: first run looked bad, second run looked normal. The main thing that moved was latency, not quality.

[visionyard23]

My first guess would be this: Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at long context. Not glamorous advice, but repeatable baselines save a lot of time.