+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Prompting & Context Engineering (https://www.datasculpting.ai/inference/forumdisplay.php?fid=15)
+--- Forum: Long Context / Memory (https://www.datasculpting.ai/inference/forumdisplay.php?fid=64)
+--- Thread: long context that stayed useful (/showthread.php?tid=566)
Pages:
1
2
long context that stayed useful - edgequill - 09-28-2026
Curious what people here are seeing.
I'm trying to decide whether to keep the current stack or change course. Main question is how people here are handling workstation thermals.
If you've tested something similar, what did you keep and what did you stop bothering with?
I care more about stability than squeezing the last possible tokens/sec out of it.
Re: long context that stayed useful - fastfield4 - 09-28-2026
A couple things I'd check first. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at structured output. Once you know the plain path works, then add the faster path back in piece by piece.
Re: long context that stayed useful - bytesupply76 - 09-28-2026
My first guess would be this: Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at workstation thermals. Not glamorous advice, but repeatable baselines save a lot of time.
Re: long context that stayed useful - phasetrack - 10-01-2026
My first guess would be this: Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at reranking. Once you know the plain path works, then add the faster path back in piece by piece.
Re: long context that stayed useful - bytesupply76 - 10-01-2026
Not sure I agree there. I get why that seems right, but on 7900 XTX the bottleneck felt elsewhere. Worth testing the simpler explanation before adopting the more interesting one.
Re: long context that stayed useful - quantghost - 10-01-2026
What exact setup are you running? Is this local-only or are you routing anything externally?
There are a few different conversations hiding inside this thread, so being explicit about the goal helps.
Re: long context that stayed useful - plainember52 - 10-04-2026
I don't think that's the whole story. I wouldn't optimize around that first. On my side the bigger swing came from the runtime and not the model. Not saying you're wrong, just that I wouldn't lock that in as the lesson yet.
Re: long context that stayed useful - flintriver50 - 10-04-2026
Thanks, that helps. I ended up in about the same place using SGLang.
Re: long context that stayed useful - torchstation - 10-04-2026
What exact setup are you running? Do you see the same thing with a smaller batch or shorter prompt?
Re: long context that stayed useful - plainember52 - 10-04-2026
Can you fill in a bit more context? Is this local-only or are you routing anything externally?