+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: LLM Models (https://www.datasculpting.ai/inference/forumdisplay.php?fid=12)
+--- Forum: Open Weight (https://www.datasculpting.ai/inference/forumdisplay.php?fid=49)
+---- Forum: Recipes & Configurations (https://www.datasculpting.ai/inference/forumdisplay.php?fid=52)
+---- Thread: routing between local and hosted with vLLM (/showthread.php?tid=542)
Pages:
1
2
Re: routing between local and hosted with vLLM - vectorshed43 - 10-02-2026
Thanks, that helps. That matches what I saw on M3 Max. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.
Re: routing between local and hosted with vLLM - torchstation - 10-05-2026
Quote:Thanks, that helps. That matches what I saw on M3 Max. I got better results once I stopped chasing one metric and just looked at whether the workflow
My first guess would be this: Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at small local stacks. Once you know the plain path works, then add the faster path back in piece by piece.
If you post your exact settings, people here can probably narrow it down pretty fast.
Re: routing between local and hosted with vLLM - llmtrack - 10-06-2026
Ran a quick check on my box. My result was mixed: first run looked bad, second run looked normal. I expected the model choice to matter more than it did.
Re: routing between local and hosted with vLLM - relaynest82 - 10-06-2026
Thanks, that helps. The pattern around speculative decoding has been pretty similar for me. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.