+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Harnesses, Agents & Orchestration (https://www.datasculpting.ai/inference/forumdisplay.php?fid=11)
+--- Forum: Agents (https://www.datasculpting.ai/inference/forumdisplay.php?fid=44)
+--- Thread: student models for local use for Phi (/showthread.php?tid=922)
student models for local use for Phi - plainworks - 10-08-2026
This keeps coming up in conversation so I figured I'd start a thread.
It feels like the conversation has shifted hard toward inference and serving details lately, and honestly that makes sense once you start dealing with small local stacks.
Maybe obvious to some of you, but it clicked for me this week.
Re: student models for local use for Phi - oxideworks59 - 10-09-2026
Quote:This keeps coming up in conversation so I figured I'd start a thread.
I don't think that's the whole story. The reported numbers can look compelling, but in ordinary usage I didn't see the same benefit. Not saying you're wrong, just that I wouldn't lock that in as the lesson yet.
Re: student models for local use for Phi - emberpeak - 10-09-2026
Did a small comparison this afternoon. I couldn't reproduce the slowdown until I pushed context much higher.
Re: student models for local use for Phi - quietstack - 10-09-2026
My first guess would be this: If you're on TensorRT-LLM, I'd confirm the version first and then reduce variables until the behavior is obvious. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it. At this point the real bottleneck is my patience.
Re: student models for local use for Phi - oxidemixer - 10-09-2026
What worked for me was pretty simple. If you're on PyTorch, I'd confirm the version first and then reduce variables until the behavior is obvious. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.
Re: student models for local use for Phi - rocmmixer - 10-09-2026
I'd approach it in a boring order. Before rebuilding anything, I'd test one clean baseline with DeepSeek and compare from there. Once you know the plain path works, then add the faster path back in piece by piece.