Stars
Simple, Lightweight, Fast and opinionated immutable Linux for desktops, servers and AI agents. A/B upgrades, encrypted disks, built from source on Hadron.
kubectl plugin that right-sizes Kubernetes workloads from CPU pressure (PSI), not just usage. Read-only, uses the Prometheus you already have, explains every number.
a community oriented 1:1, vLLM-alike (Continuous batching, paged KV) engine in C++ with additional features (GGUF, RadixAttention, Cache-aware scheduling, ...)
A local audio transcription tool that acts as a "voice keyboard" on Linux/Wayland
moss-transcribe.cpp is a from-scratch C++17 inference port of OpenMOSS MOSS-Transcribe-Diarize, built on ggml
face-detect.cpp is a from-scratch C++17 port of the insightface buffalo face-recognition pipeline, built on ggml.
ggml port of CED SOTA-tier audio-tagging model that classifies everyday sounds
A from-scratch C++17/ggml port of Depth Anything 2 and 3 (ByteDance)
A tiny, zero-dependency LLM agent harness that lives in your terminal. One static binary. No runtime, no daemon, no cloud. Local-LLM friendly. Summon it anywhere with Ctrl+Space.
OpenAI's privacy filter NER model architecture implemented in a minimal C++/GGML runtime
Port of Nvidia LocateAnything-3B on ggml
Fast and portable Parakeet implementation in C++ with ggml
Adaptive Precision for EXpert Models: MoE-aware mixed-precision quantization
Open-source models that turn Italian audio into text, locally, on your own hardware, with optional speaker diarization. Brought to you by the LocalAI team.
C++ port of Microsoft VibeVoice built on ggml
Dimensional is the agentic operating system for physical space. Command humanoids, quadrupeds, drones, and other hardware platforms in natural language and build multi-agent systems that work seaml…
Hadron is a minimal, from-scratch Linux distro using vanilla components, engineered for trusted and flexible boot environments.
Go library for building intelligent, co-operative agentic software and LLM-powered workflows, focusing on improving results for small, open source language models that scales to any LLM.
⚙️ A macOS DMG package builder with a native GUI and CLI support
RF-DETR is a real-time object detection and segmentation model architecture developed by Roboflow, SOTA on COCO, designed for fine-tuning. [ICLR 2026]
PyTorch code and models for VJEPA2 self-supervised learning from video.
🎤 A voice assistant that let's you control any Linux desktop and transcribe any audio.
Official PyTorch implementation of One-Minute Video Generation with Test-Time Training
🧠 100% Local Memory layer and Knowledge base for agents with WebUI