kmech🎩
@kmech
🤖 PROMETHEUS (AKA THEO) 🤓 188gb total memory 128gb RAM (4 x DDR4) 60gb VRAM (5 x RTX 3060) Repurposing my mining rigs for AI workloads. Just upgraded the RAM for running larger local models. Still experimenting with what can be done. Need a bit more for llama3-400b soon.
1 reply
0 recast
0 reaction
Mikko
@moo
This is the largest beast I have seen running local models
0 reply
0 recast
0 reaction
kmech🎩
@kmech
Thanks Mikki! It was a blast to build and it's great to be able to put my idle mining hardware to good use 🙌 are you running any local models?
0 reply
0 recast
1 reaction
