Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities

Official implementation for BitVLA: 1-bit Vision-Language-Action Models for Robotics Manipulation

Running Microsoft's BitNet via Electron, React & Astro

Implementing the BitNet model in Rust

Running Microsoft's BitNet inference framework via FastAPI, Uvicorn and Docker.

BitNet.Js - A node.js implementation of the microsoft bitnet.cpp inference framework.

All the important tools for BITNET/HNET programs like relay chat, xyzzy, YWAKEUP, IUCVTRAP, CLRSCRN

pure pytorch implementation of Microsoft's BitNet b1.58 2B4T

BitNet: Learning-Based-Bit-Depth-Expansion

Rust bindings for bitnet.cpp based on llama-cpp-4

Learning Path: RISC-V & Advanced Edge AI on SiFive FE310-G002 SoC | 32-bit RISC-V | 320 MHz | 16KB L1 Instruction Cache | 128Mbit (16MB) QSPI Flash | 4-stage pipeline

An unofficial implementation of BitNet

Pure Rust engine for BitNet LLMs — Conversion, Inference, Training and Research. With streaming and GPU/CPU support

Official implementation of BitMamba-2. A scalable 1.58-bit State Space Model (Mamba-2 + BitNet) trained from scratch on 150B tokens. Includes JAX training code and high-performance C++ inference engine.

Long term project about a custom AI architecture. Consist of cutting-edge technique in machine learning such as Flash-Attention, Group-Query-Attention, ZeRO-Infinity, BitNet, etc.

Transformer Bitnet en Verilog

Distily: Language Model Distillation Toolkit and Library

Ultra-lightweight C++ inference engine for BitMamba-2 (1.58-bit SSM). Runs 1B models on consumer CPUs at 50+ tok/s using <700MB RAM. No heavy dependencies.

Zero-dependency implementation of BitNet neural network training and BPE tokenization in C

This is the repo for the MixKABRN Neural Network (Mixture of Kolmogorov-Arnold Bit Retentive Networks), and an attempt at first adapting it for training on text, and later adjust it for other modalities.