because some retard breeder can't be bothered to care about their child's mental health

>>513497344 (OP)

This is part of why I run local LLMs on my laptop with KoboldCPP.

Easy peasy,
nobody sees me,
it makes the kikes and glowies uneasy.

4GB Nvidia card and 32GB memory and I can easily do decent-sized multi-model models.

>>513497344 (OP)
Why do you dorky dumbass dweebs keep talking to gAi?
Go fishing fool.

>>513497344 (OP)
So, ai bad. Parent, bad. But let me guess nigger, you could have saved her? Lol lmao

>>513497609
Do you know models I can run on Radeon?

>>513498035

Same exact models as on an NVidia. Can even run it ALL on your CPU & regular ram, it's just slower. You can do that with KoboldCPP if I recall correctly, When you launch KoboldCPP just set the "Backed:" to "Use Vulkan" for Radeon or "Use CPU" if that doesn't work. I've totally loaded 4B multimodal models and entirely run them from regular ram because the speed difference for a model that small isn't a terribly noticeable difference between VRAM and RAM.

>>513497344 (OP)
>because the powers that be hate you and want you dead
ftfy

>>513498471
Whats the best model i can run on a 5090 and 64gb of ram

>>513499071

ask on /r/LocalLLaMA/ or a Local Models General thread like >>>/g/106316518