KV cache compression worth it for Mistral

[batchscope72]

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

Short version: SGLang behaved well for me, but only after I stopped ignoring NVFP4.

Curious if anyone else landed in the same place.

[mangofield33]

Need a little more detail here. Is this local-only or are you routing anything externally?

[southden]

Good point. That matches what I saw on dual 3090. The biggest change for me was simplifying the stack instead of adding one more optimization layer.

[fusedtinker]

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.

[routecrate]

Appreciate the detail. I ended up in about the same place using MLX.

[amberrunner]

same here on my side.

[dockbrook21]

same here on my side.

Interesting, I hadn't tried that. The pattern around batch size has been pretty similar for me. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.

[mellowsupply]

worth checking the driver first.

[oxidegrove40]

Ran a quick check on my box. The issue went away once I changed one setting and restarted the stack cleanly. The main thing that moved was latency, not quality.

[kernelghost]

Interesting, I hadn't tried that. I ended up in about the same place using SGLang. The biggest change for me was simplifying the stack instead of adding one more optimization layer. The logs were somehow calmer than I was.