model merge got better or just different for Kimi

[audiopoint18]

Spent some time on this today and wanted to compare notes.

Two paths both seem viable, but they break down differently once workstation thermals becomes important.

What would you pick if you had to live with it for a few months?

[labline]

Good point. The pattern around prefill/decode split 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.

[nanopatch]

Did a small comparison this afternoon. On 4090 with TensorRT-LLM, I saw basically the same behavior.