vLLM vs SGLang for small teams

[amberrunner]

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

Short version: SGLang behaved well for me, but only after I stopped ignoring NVFP4.

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

I care more about stability than squeezing the last possible tokens/sec out of it.

If I had to summarize the lesson so far, it would be that clean setup and observability beat clever guesses.

[rocmrunner40]

Need a little more detail here. Do you see the same thing with a smaller batch or shorter prompt?

[cacheline]

My first guess would be this: Before rebuilding anything, I'd test one clean baseline with Phi and compare from there. Once you know the plain path works, then add the faster path back in piece by piece.

[cudacrate]

I'd start with the simpler stack.

[southspark]

My first guess would be this: On B200, the easiest win is often to back off the aggressive settings and see whether the problem survives.

[amberrunner]

A couple things I'd check first. On dual 3090, the easiest win is often to back off the aggressive settings and see whether the problem survives. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.

There are a few different conversations hiding inside this thread, so being explicit about the goal helps.

[thetaplane]

A couple things I'd check first. On dual 3090, the easiest win is often to back off the aggressive settings and see whether the problem survives. A lo

Good point. The pattern around reranking has been pretty similar for me. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.

[deltapoint17]

Appreciate the detail. The pattern around prefix caching has been pretty similar for me.

[mangofield33]

Thanks, that helps. The pattern around structured output has been pretty similar for me. The biggest change for me was simplifying the stack instead of adding one more optimization layer.

[edgequill]

What exact setup are you running? What model, what runtime, and roughly what context size?