Run a 4B-parameter LLM in 942 MiB — PyTorch-native BitNet b1.58 with zero float multiplications at inference
Simonc44/Ternair: Run a 4B-parameter LLM in 942 MiB — PyTorch-native BitNet b1.58 with zero float multiplications at inference
Related
super-bison-32/pixverse-desktop: PixVerse 2026 desktop setup with full format suite, multitrack timeline and portable mode for Windows 11
PixVerse 2026 desktop setup with full format suite, multitrack timeline and portable mode for Windows 11
devin667/comfyui-dsh: ComfyUI 本地生图 + DSH 接入：一键安装、Denoise 滑杆页面、workflow 配置。不经云端内容审核，出图全在本机。
ComfyUI 本地生图 + DSH 接入：一键安装、Denoise 滑杆页面、workflow 配置。不经云端内容审核，出图全在本机。
alicebrown323/pro-tools-desktop: Pro Tools 2026 desktop setup with multitrack editing, MIDI sequencing, VST hosting and a portable mode for Windows 11
Pro Tools 2026 desktop setup with multitrack editing, MIDI sequencing, VST hosting and a portable mode for Windows 11