context size vs speed in GGUF: setup advice

[northcraft90]

What worked for me was pretty simple. Before rebuilding anything, I'd test one clean baseline with Gemma and compare from there. Not glamorous advice, but repeatable baselines save a lot of time.

[batchpeak]

same here on my side.

[vrambandit]

I'd start with the simpler stack.

[pixelcrate82]

Good point. That matches what I saw on M3 Max. The biggest change for me was simplifying the stack instead of adding one more optimization layer.