eicker@lemmy.world to Technology@lemmy.worldEnglish · 8 days agoSam Altman Absolutely Hates the Idea of People Running Their Own AI Models at Home, Argues They Shouldn’t Own Computers: He'd prefer if we surrender all our RAM to data centers.futurism.comexternal-linkmessage-square365linkfedilinkarrow-up11.38Karrow-down118
arrow-up11.36Karrow-down1external-linkSam Altman Absolutely Hates the Idea of People Running Their Own AI Models at Home, Argues They Shouldn’t Own Computers: He'd prefer if we surrender all our RAM to data centers.futurism.comeicker@lemmy.world to Technology@lemmy.worldEnglish · 8 days agomessage-square365linkfedilink
minus-squareSirMaple__@lemmy.calinkfedilinkEnglisharrow-up6arrow-down1·8 days agoI’m glad I have 2 ASUS GX10 devices connected together. Runs DeepSeek 0731 beautifully at 128K context even. I also have 2 M4 Pro with 24GB each that run smaller utility LLM’s.
minus-squarepigup@lemmy.worldlinkfedilinkEnglisharrow-up4arrow-down1·8 days agoBro, 0731 is a hallucinating mess. Premiere model is https://mia-ai.net/models/GLM-5.3-Flash-EXL3-2x-DGX-Sparks-TensorFold I can vouch for its greatness. Haven’t touched my Claude since I got it running.
minus-squareTiredDinoByte@lemmy.todaylinkfedilinkEnglisharrow-up3·8 days agoI’ll have to try that, for my part qwen-3.8-flash-next has been a world of difference
minus-squarepigup@lemmy.worldlinkfedilinkEnglisharrow-up3·8 days agoThat’s a great model for sure, I run it on my strix halo
minus-squareSirMaple__@lemmy.calinkfedilinkEnglisharrow-up2·8 days agoI’ve been meaning to try the GLM line on my GX10 pair. Maybe this weekend I’ll see about getting it going. Tho the 0731 works pretty well with SillyTavern and is not half bad with yaml and ansible playbooks. But I will be giving the GLM a spin and see how it is.
minus-squarepigup@lemmy.worldlinkfedilinkEnglisharrow-up1·8 days agoI use GLM as my network admin. 0731 made a mess of my home lab, GLM is frontier quality. Rarely makes mistakes.
I’m glad I have 2 ASUS GX10 devices connected together. Runs DeepSeek 0731 beautifully at 128K context even.
I also have 2 M4 Pro with 24GB each that run smaller utility LLM’s.
Bro, 0731 is a hallucinating mess. Premiere model is https://mia-ai.net/models/GLM-5.3-Flash-EXL3-2x-DGX-Sparks-TensorFold
I can vouch for its greatness. Haven’t touched my Claude since I got it running.
I’ll have to try that, for my part qwen-3.8-flash-next has been a world of difference
That’s a great model for sure, I run it on my strix halo
I’ve been meaning to try the GLM line on my GX10 pair. Maybe this weekend I’ll see about getting it going.
Tho the 0731 works pretty well with SillyTavern and is not half bad with yaml and ansible playbooks.
But I will be giving the GLM a spin and see how it is.
I use GLM as my network admin. 0731 made a mess of my home lab, GLM is frontier quality. Rarely makes mistakes.