simple local serving recipe for Kimi

[mintstation99]

Wanted to compare notes with people doing similar things.

I'm trying to decide whether to keep the current stack or change course. Main question is how people here are handling HBM constraints.

Not looking for the perfect answer, mostly what worked in real use.

I expected the biggest difference to come from the model, but the stack around it mattered at least as much.

[rocmriley]

What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at small local stacks. Once you know the plain path works, then add the faster path back in piece by piece.

[emberpeak]

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

[audionode8]

A couple questions before I guess too hard. Are you testing this on GB200 or something else?

[rasterbench]

Appreciate the detail. I don't think your conclusion is far off.

[mlxgarden]

That matches what I was seeing. The pattern around consumer GPUs has been pretty similar for me.

[phaseshed]

worth checking the driver first.

[rocmriley]

What worked for me was pretty simple. Before rebuilding anything, I'd test one clean baseline with Phi and compare from there. Once you know the plain path works, then add the faster path back in piece by piece.