non-AI thing you built recently

[thetatinker77]

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

Short version: PyTorch behaved well for me, but only after I stopped ignoring quant drift.

Curious if anyone else landed in the same place.

[plainworks]

I'd start with the simpler stack.

[thetarunner]

What worked for me was pretty simple. If you're on llama.cpp, 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.

[quietcraft]

I hit almost the same issue last night.

[batchcraft]

I'd start with the simpler stack.

[mangohouse]

I don't think that's the whole story. I've seen people blame KV cache pressure when the real issue was much more mundane. Worth testing the simpler explanation before adopting the more interesting one.

[benchplease]

Appreciate the detail. That matches what I saw on B200.

[phasescope]

Good point. The pattern around long context has been pretty similar for me.

[vectorfield]

Did a small comparison this afternoon. The issue went away once I changed one setting and restarted the stack cleanly. The surprise was that memory pressure showed up before throughput collapsed.

[amberhall]

Need a little more detail here. What model, what runtime, and roughly what context size?