LM Studio surprisingly handy

[cacheworks]

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

After a few runs on GB200, I ended up with decent performance, but the setup only really made sense once I adjusted around workstation thermals.

Nothing scientific here, just practical notes from a real setup.

[oxidemixer]

I'd approach it in a boring order. If you're on Ollama, 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.

[ggufwire22]

Good point. The pattern around embeddings has been pretty similar for me. The biggest change for me was simplifying the stack instead of adding one more optimization layer.

[prompttinker44]

Good point. The pattern around embeddings has been pretty similar for me. The biggest change for me was simplifying the stack instead of adding one mo

I tested this tonight. On mini PC with vLLM, I saw basically the same behavior. I expected the model choice to matter more than it did.

[vectorcrate98]

[quote]Good point. The pattern around embeddings has been pretty similar for me. The biggest change for me was simplifying the stack instead of adding

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

[fusedhall]

[quote][quote]Good point. The pattern around embeddings has been pretty similar for me. The biggest change for me was simplifying the stack instead of

I'd approach it in a boring order. On Threadripper box, the easiest win is often to back off the aggressive settings and see whether the problem survives. Not glamorous advice, but repeatable baselines save a lot of time.

[routerunner]

What worked for me was pretty simple. Before rebuilding anything, I'd test one clean baseline with Phi and compare from there. Not glamorous advice, but repeatable baselines save a lot of time.

[vectorcrate98]

A couple questions before I guess too hard. What model, what runtime, and roughly what context size?

[vectorshed43]

Tried to reproduce it just now. The issue went away once I changed one setting and restarted the stack cleanly. The main thing that moved was latency, not quality.

[latentscope35]

I tested this tonight. The issue went away once I changed one setting and restarted the stack cleanly. The main thing that moved was latency, not quality.