vLLM vs SGLang for small teams with PyTorch

[phasestream53]

I figured I'd ask here before changing everything.

My current setup is LM Studio with A6000, and I keep running into NVFP4.

If you've tested something similar, what did you keep and what did you stop bothering with?

[signalpeak]

Good point. That matches what I saw on dual 3090. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.