+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Optimization & Performance (https://www.datasculpting.ai/inference/forumdisplay.php?fid=10)
+--- Forum: Benchmarks (https://www.datasculpting.ai/inference/forumdisplay.php?fid=43)
+--- Thread: domain tune without wrecking base behavior on B200 (/showthread.php?tid=807)
Pages:
1
2
Re: domain tune without wrecking base behavior on B200 - benchplease - 10-16-2026
Did a small comparison this afternoon. My result was mixed: first run looked bad, second run looked normal.
Re: domain tune without wrecking base behavior on B200 - turbosupply83 - 10-17-2026
same here on my side.
Re: domain tune without wrecking base behavior on B200 - cudascope - 10-17-2026
I'd approach it in a boring order. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at tokens/sec. Once you know the plain path works, then add the faster path back in piece by piece.
Re: domain tune without wrecking base behavior on B200 - vrampilot - 10-18-2026
Maybe, but I had the opposite result. 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.
Re: domain tune without wrecking base behavior on B200 - circuitworks90 - 10-18-2026
Tried to reproduce it just now. I couldn't reproduce the slowdown until I pushed context much higher. The main thing that moved was latency, not quality.
Re: domain tune without wrecking base behavior on B200 - cacheline - 10-18-2026
What worked for me was pretty simple. Before rebuilding anything, I'd test one clean baseline with Kimi and compare from there. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.