+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Inference Engines (https://www.datasculpting.ai/inference/forumdisplay.php?fid=9)
+--- Forum: vLLM / SGLang (https://www.datasculpting.ai/inference/forumdisplay.php?fid=36)
+--- Thread: continuous batching wins and losses (/showthread.php?tid=238)
Pages:
1
2
Re: continuous batching wins and losses - tensorkite - 09-20-2026
What worked for me was pretty simple. On GB200, the easiest win is often to back off the aggressive settings and see whether the problem survives.
Re: continuous batching wins and losses - deltapoint17 - 09-20-2026
I'd start with the simpler stack.
Re: continuous batching wins and losses - routearch - 09-20-2026
Appreciate the detail. I don't think your conclusion is far off. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.