With 128GB of ram on a Mac, GLM 4.5 Air is going to be one of your best options. You could run it anywhere from Q5 to Q8 depending on how you wanna manage your speed to quality ratio.
I have a different system that likely runs it slower than yours will, and I get 5 T/s generation which is just about the speed I read at. (Using q8)
I do hear that ollama may be having issues with that model though, so you may have to wait for an update to it.
I use llamacpp and llama-swap with openwebui, so if you want any tips on switching over I'd be happy to help. Llamacpp is usually one of the first projects to start supporting new models when they come out.
Edit: just reread your post. I was thinking it was a newer Mac lol. This may be a slow model for you, but I do think it'll be one of the best your can run.
Yeah I've given up on making a living off of art. I've pivoted to trying to just share it for free as much as I can. Honestly I think big corporate studios would have way less of a market if community art was more of a thing. It'd be nice to have local art and culture around rather than soulless corpo slop. I think those communities would process the world in a healthier way.
I hear you though. It really really sucks ass that literally no one (if you round) gets to make a living doing art. And (if you don't round) those that do get to make a living on it generally come from a very small subset of the available cultures.
This is a good thing to be mad about. It affects how everyone sees the world.
I think you are underestimating how adaptable humans are. We absolutely conform to the systems that govern us, and they are NOT equally likely to produce bad outcomes.
The only place I see those terms used to describe benign behavior is in rage bait. My guess is that you haven't heard someone use those terms in a serious way because of an internet bubble or something.
I bet you could find info on the rest pretty easily.
Also its real ironic that your comment meets the definition of mansplaining lol.
"(for a man) to comment on or explain something, to a woman, in a condescending, overconfident, and often inaccurate or oversimplified manner".[1][2][3][4]
I'm not a woman though, but if any read your comment then you meet the strict definition lol.
Ah fair. I was confused by the language. I thought you were saying broadly that AI was bad because it is unnatural rather than because it is cultural / technological.
I still don't think you've made a good case though. We have lots of tech that had us think less that didn't lead to the end of humanity. What would make AI special enough to end us? People have felt apocalyptic about a lot of things.
From my perspective the issues aren't about AI itself. The issue is that a small group of folks control our society and how tech gets used.
I am here to call out the natural/unnatural fallocy. It is silly. Can you really draw a line between the two (natural vs unnatural) in a way that is logical and still supports your argument?
Around me sativa/indica is just a short hand for how sleepy weed is. No one (around me) pays attention if the weed is actually genetically close to either.
I think its just habit that I use those words rather than saying sleepy or energizing.
You're gonna have to defend the socialism take with this audience (including me). From my perspective you disregarded a huge set of possible solutions then said "idk how to solve this".
In the other comment you said you haven't done much research. There is a lot of diversity in leftism. I personally lean towards anarcho federalism.
Nothing is truly hopeless. The world is far too complicated to seal any future shut all the way. There will always be routes. (Though it may end up as unlikely as your perfect world materializing instantly due to quantum teleportation)
With 128GB of ram on a Mac, GLM 4.5 Air is going to be one of your best options. You could run it anywhere from Q5 to Q8 depending on how you wanna manage your speed to quality ratio.
I have a different system that likely runs it slower than yours will, and I get 5 T/s generation which is just about the speed I read at. (Using q8)
I do hear that ollama may be having issues with that model though, so you may have to wait for an update to it.
I use llamacpp and llama-swap with openwebui, so if you want any tips on switching over I'd be happy to help. Llamacpp is usually one of the first projects to start supporting new models when they come out.
Edit: just reread your post. I was thinking it was a newer Mac lol. This may be a slow model for you, but I do think it'll be one of the best your can run.