@HuggingModels The community is really hyped about local self-hosted LLMs, but thereโs still a lack of clarity. So, Iโd love to know from the communityโwhat should I do? What setup would give me the best performance without having to constantly expand?
@HuggingModels the 13 billion ones, they either ran slowly poorly, or didnโt respond accurately. So, I created a cluster with two Mac Minis, 16 GB each, using a tool called Exo, which lets you cluster them and run models But itโs still slow. Exo works with OpenCode, but itโs really sluggish..>
@humzaakhalid I think the Mac hype is overrated. No one is being specific about what you can run on it, what kind of hardware you need, what the available models are to run easily on a Mac Mini, and which Mac you would need. We all know not everyone can run the 120B+models if you don't have HW
@TheDemSlayer@POTUS I wanted to ask why the fuck your boys on the other side of the world and not home then i saw the Israeli dirty flaq then i released no need for that question
@usembthehague Lol, you guys are never happy with anyone. Just don't do like you did with Iraq. After a good partnership with the government, you guys started to not like it, then boom, boom.