comparing models without fooling yourself: what are you seeing?

[cudascope]

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

I tried this with Phi using LM Studio on Threadripper box. The headline result was better than I expected, but consumer GPUs still ended up being the limiting factor.

Happy to post more details if anyone wants them.

My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.

[batchsize]

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.

There are a few different conversations hiding inside this thread, so being explicit about the goal helps.

[latentcraft]

What worked for me was pretty simple. If you're on SGLang, 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.

[audionode8]

What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at workstation thermals.

[llm_wren]

Interesting, I hadn't tried that. I don't think your conclusion is far off. The biggest change for me was simplifying the stack instead of adding one more optimization layer.

[benchplease]

Need a little more detail here. Are you testing this on 5090 or something else?

[thetapath]

My first guess would be this: On 7900 XTX, the easiest win is often to back off the aggressive settings and see whether the problem survives.

[nanobrook]

Good point. I don't think your conclusion is far off. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.

[cacheline]

What exact setup are you running? Is this local-only or are you routing anything externally?