ok heres a break down of my specs
9950x3d with 96gb of 6400 mhz ram
a 7900 xtx
a 6800 (pcie 4.0 4x)
a 6600 (pcie 4.0 4x chipset)
and i am running gentoo linux and my front end is going to be silly tavern
how should i best configure this
i want to run something with the highest detail and best rpesonses as possible but also quick responses, with more of a focus on the image gen the the llm.
how should i split up which moddel on which gpu could i use one of the gpus for pre prossesing the info befor handing it off to the next one and should i use my cpu/igpu sense i have alot of ram and is it worth it to share the ram on my cpu with something like my 7900 xtx or would that be too much of a preformance hit.
i got no idea what i am doing as i have not messed around with local ai befor
oh yeah also i would like to see if i could train a image gen moddel for fun recommendations o how to do that as well would be nice