long context that stayed useful: anyone testing this?

[nanobrook]

This keeps coming up in conversation so I figured I'd start a thread.

The more I work with local setups, the more it seems like continuous batching decides everything downstream.

Wondering if this is just where the whole space is heading now.

For context, I'm not chasing leaderboard numbers. I mostly want the system to feel consistent day to day.

[quietarch]

subscribed, curious where this lands.

[tinycontext]

subscribed, curious where this lands.

That matches what I was seeing. That matches what I saw on Threadripper box. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.

I'm happy to test one more variant if you want a second data point.

[torchkite]

I tested this tonight. I couldn't reproduce the slowdown until I pushed context much higher. The surprise was that memory pressure showed up before throughput collapsed.

[ggufvault]

Maybe, but I had the opposite result. I get why that seems right, but on B200 the bottleneck felt elsewhere. Could be a workload difference, but I wouldn't generalize too fast from one setup.

[fusedhall]

A couple things I'd check first. Before rebuilding anything, I'd test one clean baseline with FLUX and compare from there. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it. Turns out 'just one more test' was a lie.

[emberquill]

Maybe, but I had the opposite result. I wouldn't optimize around that first. On my side the bigger swing came from the runtime and not the model.

[tinycontext]

Did a small comparison this afternoon. On 5090 with PyTorch, I saw basically the same behavior. The main thing that moved was latency, not quality.

[turboyard94]

That matches what I was seeing. I don't think your conclusion is far off. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.

[quantghost]

What worked for me was pretty simple. On 4090, 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.