cm0002@ttrpg.network to Technology@lemmy.zipEnglish · 2 months agoMicrosoft just open-sourced bitnet.cpp, a 1-bit LLM inference framework. It let's you run 100B parameter models on your local CPU without GPUs. 6.17x faster inference and 82.2% less energy on CPUs.github.comexternal-linkmessage-square0linkfedilinkarrow-up11arrow-down10cross-posted to: technology@lemmy.mllocalllama@sh.itjust.works
arrow-up11arrow-down1external-linkMicrosoft just open-sourced bitnet.cpp, a 1-bit LLM inference framework. It let's you run 100B parameter models on your local CPU without GPUs. 6.17x faster inference and 82.2% less energy on CPUs.github.comcm0002@ttrpg.network to Technology@lemmy.zipEnglish · 2 months agomessage-square0linkfedilinkcross-posted to: technology@lemmy.mllocalllama@sh.itjust.works
