prompt length getting silly

[thetaarch43]

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 structured output.

Maybe obvious to some of you, but it clicked for me this week.

[cozywire]

Thanks, that helps. I don't think your conclusion is far off.

[mintstation99]

What worked for me was pretty simple. If you're on LM Studio, I'd confirm the version first and then reduce variables until the behavior is obvious. Not glamorous advice, but repeatable baselines save a lot of time.

[nanopatch]

Need a little more detail here. Are you testing this on Threadripper box or something else?

[mlxgarden]

Thanks, that helps. I don't think your conclusion is far off.

[dockember]

What exact setup are you running? Do you see the same thing with a smaller batch or shorter prompt?

[thetaarch43]

What worked for me was pretty simple. If you're on vLLM, I'd confirm the version first and then reduce variables until the behavior is obvious.

[promptspark58]

What worked for me was pretty simple. If you're on vLLM, I'd confirm the version first and then reduce variables until the behavior is obvious.

Ran a quick check on my box. I couldn't reproduce the slowdown until I pushed context much higher. The main thing that moved was latency, not quality.

[torchtrack24]

Did a small comparison this afternoon. On M3 Max with SGLang, I saw basically the same behavior.

[tensorkite]

Thanks, that helps. That matches what I saw on A6000.