bitnet-b158

TQ1.0 ternary inference engine for BitNet b1.58 on CPU. Pack + run Falcon3-1B/3B/7B/10B, no GPU needed.

Dual‑Helix Resonance Neural Architecture

Official inference framework for 1-bit LLMs

First 1-bit (BitNet b1.58) recursive reasoner for Sudoku-Extreme - distilled from a 7M-param FP TRM teacher into a 1.4 MB ternary student

A lightweight Ternary LLM (BitNet b1.58)

BitQuant allows you to apply quantization (with Straight-Through Estimator support) to your existing PyTorch models without altering the internal layer implementations

Dependency-light Rust runtime for BitNet b1.58-style LLMs