OSRepos

PEFT: State-of-the-Art Parameter-Efficient Fine-Tuning

PEFT (Parameter-Efficient Fine-Tuning) is a cutting-edge library from Hugging Face designed to efficiently adapt large pretrained models for various downstream applications. It dramatically reduces computational and storage costs by fine-tuning only a small subset of model parameters. This approach enables achieving performance comparable to fully fine-tuned models, making advanced AI accessible on more modest hardware.

May 14, 2026

Stunner: Quickly Detect Your NAT Type with Multi-Server STUN

Stunner is an efficient Go CLI tool designed to accurately detect your Network Address Translation (NAT) type. By sending STUN Binding Requests to multiple servers, it classifies your NAT as Full Cone, Symmetric, or Restricted, providing crucial insights into your network environment. This helps users understand network behavior and assess the feasibility of techniques like hole punching.

Lychee: A Stunning and User-Friendly Photo Management System

Lychee is a free, open-source photo management system designed for self-hosting. It offers a beautiful and intuitive interface, allowing users to easily upload, organize, and share their photo collections. With quick installation and essential features, Lychee provides a secure way to manage your personal photos.

fastFM: A High-Performance Python Library for Factorization Machines

fastFM is a powerful Python library designed for Factorization Machines, offering high-performance implementations of various optimization routines. It integrates seamlessly with the scikit-learn API, making it accessible for machine learning practitioners. The library supports regression, classification, and ranking problems, leveraging C and Cython for speed-critical operations.

Firecrawl: Web Scraping and Interaction API for AI Agents

Firecrawl is an open-source API designed to empower AI agents and applications with clean, structured web data. It provides robust capabilities for searching, scraping, and interacting with the web at scale, effectively transforming complex web content into LLM-ready formats. This tool handles the intricate challenges of web data extraction, allowing developers to focus on building intelligent applications.

Cheerio: Fast and Flexible HTML/XML Parsing and Manipulation Library

Cheerio is a popular library for parsing and manipulating HTML and XML documents in Node.js. It provides a jQuery-like API, making it easy to select, traverse, and modify elements with proven syntax. Known for its blazingly fast performance and incredible flexibility, Cheerio is an excellent choice for web scraping and server-side DOM manipulation.

Docs: Collaborative Note-Taking, Wiki, and Documentation Platform

Docs is an open-source, real-time collaborative editor designed to help teams write, organize, and share knowledge efficiently. Built with Django and React, it offers a robust alternative to tools like Notion or Google Docs, emphasizing data ownership and self-hosting capabilities. It empowers public organizations, companies, and open communities to manage their knowledge collaboratively.

Scramble: An Open-Source AI Writing Assistant for Your Browser

Scramble is an open-source Chrome extension that leverages AI to enhance your writing directly in the browser. It offers a customizable alternative to Grammarly, allowing users to configure LLM providers, models, and endpoints. With support for various large language models and custom prompts, Scramble provides flexible text enhancement options.

LLMGym: A Unified Environment for LLM Agent Development and Benchmarking

LLMGym is a unified environment interface designed for developing and benchmarking LLM applications that learn from feedback. It provides a suite of seamlessly swappable environments, making fair and comprehensive comparisons easier for researchers and developers. This project aims to be the "gym" for LLM agents, offering an intuitive interface for various tasks.

AI Chat: A Modern Laravel Starter Kit for Real-time AI Applications

AI Chat is a robust Laravel starter kit designed for building modern AI-powered chat applications. It features real-time streaming responses using the Prism PHP SDK, integrated with Inertia.js, Vue.js, and TailwindCSS, providing a solid foundation for dynamic user experiences. This kit enables developers to quickly launch AI chat solutions with a scalable backend and comprehensive features.

tortoise.cpp: Local Text-to-Speech with GGML and C++

tortoise.cpp is a C++ re-implementation of the popular Tortoise-TTS model, leveraging the efficient GGML library. This project enables high-quality, local text-to-speech generation without the need for Python dependencies. It aims to make advanced speech synthesis more accessible and performant on various hardware configurations.