benchmark result I don't trust

[cacheline]

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

Short version: Ollama behaved well for me, but only after I stopped ignoring quant drift.

Curious if anyone else landed in the same place.

[routearch]

Ran a quick check on my box. On Threadripper box with MLX, I saw basically the same behavior. The main thing that moved was latency, not quality.

[kernelghost]

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. Could be a workload difference, but I wouldn't generalize too fast from one setup.

[routearch]

My first guess would be this: If you're on TensorRT-LLM, I'd confirm the version first and then reduce variables until the behavior is obvious. Once you know the plain path works, then add the faster path back in piece by piece.

[kernelmike]

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. Once you know the plain path works, then add the faster path back in piece by piece.

[cacheline]

I think the bottleneck is somewhere else. I wouldn't optimize around that first. On my side the bigger swing came from the runtime and not the model. Not saying you're wrong, just that I wouldn't lock that in as the lesson yet.

[routearch]

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

[oxideworks59]

A couple things I'd check first. Before rebuilding anything, I'd test one clean baseline with Gemma and compare from there. Once you know the plain path works, then add the faster path back in piece by piece.

[localtinker]

I hit almost the same issue last night.

[batchpeak]

My first guess would be this: If you're on LM Studio, 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.