human evals on small teams

[fusedtinker]

This keeps coming up in conversation so I figured I'd start a thread.

It feels like the conversation has shifted hard toward inference and serving details lately, and honestly that makes sense once you start dealing with KV cache pressure.

Wondering if this is just where the whole space is heading now.

My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.

[fastlab]

This keeps coming up in conversation so I figured I'd start a thread.

My first guess would be this: Before rebuilding anything, I'd test one clean baseline with Llama and compare from there. Once you know the plain path works, then add the faster path back in piece by piece.

[macrostack69]

A couple things I'd check first. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at KV cache pressure. Once you know the plain path works, then add the faster path back in piece by piece.

[lablab]

Need a little more detail here. Do you see the same thing with a smaller batch or shorter prompt?

[audioforge]

Need a little more detail here. Do you see the same thing with a smaller batch or shorter prompt?

Thanks, that helps. I ended up in about the same place using Ollama. The biggest change for me was simplifying the stack instead of adding one more optimization layer.

There are a few different conversations hiding inside this thread, so being explicit about the goal helps.

[thetaarch43]

Interesting, I hadn't tried that. That matches what I saw on dual 3090.

[routeforge64]

Interesting, I hadn't tried that. That matches what I saw on dual 3090.

[amberhall]

Not sure I agree there. I wouldn't optimize around that first. On my side the bigger swing came from the runtime and not the model. Worth testing the simpler explanation before adopting the more interesting one.

[macrostack69]

What exact setup are you running? What model, what runtime, and roughly what context size?

[cudacrate]

A couple questions before I guess too hard. Is this local-only or are you routing anything externally?