context size vs speed in GGUF for Mixtral

[griffinloom66]

Has anyone else run into this?

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

If you've tested something similar, what did you keep and what did you stop bothering with?

I care more about stability than squeezing the last possible tokens/sec out of it.

[latentcraft]

Has anyone else run into this?

Ran a quick check on my box. On 7900 XTX with MLX, I saw basically the same behavior. I expected the model choice to matter more than it did.

[vectorport]

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

[phasescope]

Ran a quick check on my box. The issue went away once I changed one setting and restarted the stack cleanly.

[sparrowhall]

Ran a quick check on my box. The issue went away once I changed one setting and restarted the stack cleanly.

I tested this tonight. On Threadripper box with PyTorch, I saw basically the same behavior.

[vrambandit]

Appreciate the detail. That matches what I saw on 7900 XTX. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.

[dockridge]

My first guess would be this: On M3 Max, the easiest win is often to back off the aggressive settings and see whether the problem survives.

[pixelcrate82]

that fixed it for me too.

[visionyard23]

that fixed it for me too.

Thanks, that helps. I don't think your conclusion is far off. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.