Improvement: Toggle to disable the new "prompt warmup" feature

[abc-nix]

I have a single slot for llama.cpp, and it is annoying that I have to reprocess the full chat every time I switch to a different conversation to copy a part of a reply. Could you add an option in the Model or Api tab to disable this new feature? If the options is already there, please guide me to where I can find it.

Thanks for your time.

[Rose22]

the prompt warmup feature has introduced like hundreds of new bugs to openlumara and i've been debugging for days. for now i've disabled it outright until it's more stable, but i have to fix a bunch of regressions to the webui too, before i publish that change. prompt warmup REALLY messed things up badly.

[abc-nix]

Great to know the issues were not related to my changes. I pulled the latest changes and it is much better, thanks.

Thanks for the efforts on keeping this project working.