+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Miscellaneous Discussion (https://www.datasculpting.ai/inference/forumdisplay.php?fid=24)
+--- Forum: General (https://www.datasculpting.ai/inference/forumdisplay.php?fid=25)
+--- Thread: what are you all running this week (/showthread.php?tid=180)
what are you all running this week - ggufwire22 - 09-07-2026
Did a bit of testing and thought I'd post the result here.
After a few runs on 4090, I ended up with decent performance, but the setup only really made sense once I adjusted around long context.
Curious if anyone else landed in the same place.
Re: what are you all running this week - rocmriley - 09-07-2026
Appreciate the detail. The pattern around speculative decoding has been pretty similar for me. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.
Re: what are you all running this week - rocmstack25 - 09-08-2026
Need a little more detail here. Do you see the same thing with a smaller batch or shorter prompt?
Re: what are you all running this week - rocmrunner40 - 09-08-2026
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. Once you know the plain path works, then add the faster path back in piece by piece.
There are a few different conversations hiding inside this thread, so being explicit about the goal helps.
Re: what are you all running this week - vectorshed43 - 09-10-2026
I'd approach it in a boring order. Before rebuilding anything, I'd test one clean baseline with Mixtral and compare from there. Once you know the plain path works, then add the faster path back in piece by piece.
Re: what are you all running this week - quiettoken - 09-11-2026
Quote:I'd approach it in a boring order. Before rebuilding anything, I'd test one clean baseline with Mixtral and compare from there. Once you know the plai
Good point. I don't think your conclusion is far off. The biggest change for me was simplifying the stack instead of adding one more optimization layer.