Repository navigation

quant-forge

Public

About

⚡ Fused naked-PTX ternary (1.58-bit) matrix multiplication engine with zero dequantization overhead in Zig

bitnet cuda-ptx deep-learning gemm gpu-kernel quantization ternary zig

0 stars

0 watching

Report repository

Name		Name	Last commit message	Last commit date
.github/workflows		.github/workflows
src		src
.gitignore		.gitignore
Dockerfile		Dockerfile
LICENSE		LICENSE
README.md		README.md
build.zig		build.zig

Repository files navigation

⚡ quant-forge: Naked PTX 1.58-Bit Ternary Matrix Engine

Direct hardware register ternary packed GEMM kernel executing in Zig and naked CUDA PTX.

About

⚡ Fused naked-PTX ternary (1.58-bit) matrix multiplication engine with zero dequantization overhead in Zig

bitnet cuda-ptx deep-learning gemm gpu-kernel quantization ternary zig

0 stars

0 watching

Report repository

Releases

Packages

Contributors

Languages