simple router policy ideas: anyone testing this?

[turboscope5]

Did a bit of testing and thought I'd post the result here.

Short version: TensorRT-LLM behaved well for me, but only after I stopped ignoring tail latency.

Curious if anyone else landed in the same place.

[mangovault67]

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

[batchpeak]

My first guess would be this: Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at token cost. Once you know the plain path works, then add the faster path back in piece by piece.

[cobaltspark]

That matches what I was seeing. 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.

If you post your exact settings, people here can probably narrow it down pretty fast.

[phasestream53]

Good point. 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.

[sanepeak48]

Good point. 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 f

A couple things I'd check first. If you're on PyTorch, I'd confirm the version first and then reduce variables until the behavior is obvious. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it. I keep promising myself this is the last rebuild.

[cudapoint]

same here on my side.

[emberpeak]

subscribed, curious where this lands.

[kernelhouse70]

A couple things I'd check first. On Threadripper box, the easiest win is often to back off the aggressive settings and see whether the problem survives. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.

[syshouse]

I tested this tonight. I couldn't reproduce the slowdown until I pushed context much higher.