Awesome GitHub RepositoriesIncremental Updates

LightRAG is a graph-based retrieval framework designed to build retrieval-augmented generation pipelines. It structures unstructured text into knowledge graphs, enabling multi-hop reasoning and complex query synthesis across large document collections. By integrating dense vector embeddings with structured knowledge graphs, the system facilitates both similarity-based and relationship-aware information retrieval. The framework distinguishes itself through a dual-level retrieval strategy that combines low-level keyword matching with high-level semantic graph traversal to capture both specific

Graphiti is a backend framework and memory server designed to provide artificial intelligence agents with persistent, time-aware knowledge graph storage. It functions as a memory layer that enables agents to maintain context across long-term interactions by recording and evolving structured data over time. The system distinguishes itself through a specialized temporal graph database that tracks how entities and relationships change using validity windows. By combining semantic vector similarity, keyword matching, and graph topology traversal, the engine performs hybrid retrieval to locate rel

This project is a static code analysis tool and local-first code indexer that builds a persistent dependency graph of functions, classes, and imports. It functions as an AI context optimizer and codebase dependency graph, designed to reduce token usage by providing AI assistants with only the most relevant code fragments and impact analysis for a given change. The system implements a Model Context Protocol server that exposes code intelligence and architectural graph queries to external AI coding tools. It distinguishes itself by computing the change blast radius and risk scores of modificati

This project is an educational resource providing practical code examples and implementations of machine learning algorithms using the Python language. It serves as a guide for constructing predictive pipelines, clustering models, and dimensionality reduction within the Scikit-Learn ecosystem. The repository includes comprehensive demonstrations for supervised and unsupervised learning, as well as detailed examples for implementing neural networks and deep architectures. It also provides practical guidance on exporting model parameters to JSON and wrapping trained models in web APIs for produ

This project is a structured learning curriculum and technical reference for mastering deep learning with TensorFlow. It provides a comprehensive guide for building, training, and deploying neural networks, combining theoretical fundamentals with practical implementation examples. The repository distinguishes itself by covering the end-to-end machine learning workflow, from low-level tensor mathematics and linear algebra to the creation of complex model architectures. It includes specific guidance on developing data pipelines for diverse data types, such as images, text, and time-series seque

This project is a manifold learning and non-linear dimensionality reduction library used to project high-dimensional data into lower-dimensional spaces while preserving topological structure. It functions as a parametric embedding framework and a topological data visualization library for identifying clusters and patterns within complex datasets. The library distinguishes itself through parametric neural mapping, which uses neural networks to learn functional mappings that allow for out-of-sample projections and the reconstruction of original data. It supports supervised and semi-supervised d

Atlas is a modularization system and dynamic component framework for Android. It functions as a class isolation layer and incremental update engine, allowing application logic and resources to be decoupled into independent bundles that are loaded at runtime. The project distinguishes itself by providing physical bundle isolation through custom class loaders to prevent dependency conflicts and bypass method count limits. It enables the deployment of remote components and incremental patches, using bytecode diffs and resource patching to update specific application parts without requiring a ful

BERTopic is a topic modeling library used to extract interpretable themes from collections of text documents and images. It functions as a document clustering framework that transforms unstructured data into numerical vectors to group semantically similar content. The project distinguishes itself through a multimodal embedding tool that allows for joint clustering of text and images in a shared vector space. It also features a class-based TF-IDF representation engine to identify representative words for clusters and an integrated system for using large language models to generate natural lang

River es un framework de Python para machine learning online, diseñado para entrenar y evaluar modelos en datos de streaming. Permite el aprendizaje incremental actualizando los parámetros del modelo una observación a la vez, eliminando la necesidad de almacenar datasets de entrenamiento completos en memoria. La librería se distingue por un sistema dedicado de detección de concept drift que monitorea cambios en las distribuciones de datos para disparar la adaptación del modelo. También proporciona un framework de validación progresiva que simula el despliegue en tiempo real probando modelos en muestras antes de usarlos para el entrenamiento. El sistema cubre un amplio rango de capacidades de streaming, incluyendo ingeniería de características en tiempo real, pronóstico de series temporales y detección de anomalías online. Soporta aprendizaje no supervisado mediante clustering incremental y árboles de decisión, así como agregación de ensamblajes y políticas de bandidos para la selección de modelos. El proyecto incluye utilidades para la ingesta de datos de streaming desde fuentes como archivos CSV y APIs, así como herramientas para calcular estadísticas en ejecución y sketches de datos eficientes en memoria.

QOwnNotes is a desktop note editor that stores each note as a plain-text Markdown file on the local filesystem, avoiding proprietary formats and enabling direct file access. It functions as a Nextcloud Notes client, syncing notes and metadata with Nextcloud or ownCloud servers through a companion API service for versioning and sharing. The application also integrates with AI providers and exposes a local MCP server for external agents to search and fetch notes, and includes a companion browser extension for capturing web content, bookmarks, and screenshots. The editor distinguishes itself thr

GluonTS es una librería de series temporales probabilísticas y framework de pronóstico de aprendizaje profundo. Proporciona un kit de herramientas para construir, entrenar y evaluar arquitecturas de redes neuronales que predicen valores futuros como distribuciones de probabilidad para cuantificar la incertidumbre. El proyecto se distingue por soportar el pronóstico zero-shot e integrar diversos enfoques de modelado, incluyendo redes neuronales probabilísticas profundas y envoltorios para librerías estadísticas externas como Prophet y R forecast. Implementa primitivas arquitectónicas especializadas como convoluciones causales y redes residuales invertibles para prevenir la fuga de información y mapear representaciones latentes en distribuciones de probabilidad válidas. El framework cubre una superficie de ingeniería de datos integral, incluyendo escalado de series temporales, transformaciones biyectivas y modelado jerárquico. Utiliza Apache Arrow y Parquet para el streaming de conjuntos de datos de alto rendimiento y la gestión de acceso aleatorio. Para la evaluación de modelos, incluye una suite de evaluación para medir la precisión del pronóstico y la cobertura probabilística utilizando métricas como la pérdida de cuantiles y puntuaciones de probabilidad de rango continuo. La librería soporta el despliegue de modelos a través de la integración con Amazon SageMaker.

GluonTS es un framework para el pronóstico probabilístico de series temporales, diseñado para predecir valores futuros como distribuciones de probabilidad con intervalos de confianza. Soporta tanto el entrenamiento de modelos tradicionales como el pronóstico zero-shot, donde modelos preentrenados generan predicciones para nuevas series sin entrenamiento adicional. El proyecto se distingue por integrar una amplia variedad de enfoques de pronóstico en un flujo de trabajo unificado. Esto incluye arquitecturas de aprendizaje profundo como redes neuronales recurrentes y convoluciones causales, así como la integración de modelos estadísticos externos, la librería Prophet y paquetes de R. El kit de herramientas proporciona una superficie integral para la ingeniería de datos de series temporales, cubriendo el escalado de conjuntos de datos, la división y la transformación de datos temporales sin procesar en tensores. También incluye un conjunto de herramientas de evaluación para medir la precisión del pronóstico y los intervalos de incertidumbre, así como utilidades para la persistencia de conjuntos de datos utilizando formatos como Arrow y Parquet. El framework soporta el despliegue de modelos de pronóstico dentro de la infraestructura en la nube.

Robust es un framework de hotfix para Android y actualizador de aplicaciones en tiempo de ejecución diseñado para parchear errores en entornos de producción. Funciona como un parcheador de bytecode y herramienta de inyección de código dinámico, permitiendo la modificación de la lógica de la aplicación sin requerir una actualización completa de la app o un reinicio del dispositivo. El proyecto permite la remediación de defectos en Android inyectando nuevas clases y métodos en builds existentes y modificando las implementaciones de métodos existentes. Utiliza una cadena de herramientas para generar archivos de parche binarios que actualizan el comportamiento de la aplicación e introducen soluciones alternativas sobre la marcha. El sistema logra estas actualizaciones mediante redirección de bytecode, aislamiento de cargador de clases y parcheo de tablas de métodos en tiempo de ejecución. También incorpora hooking basado en reflexión e inyección dinámica de bytecode para redirigir el flujo de ejecución dentro de la máquina virtual en ejecución.

This project is a distributed machine learning platform and sparse deep learning framework designed for training and serving models with high-dimensional sparse data. It functions as an online model serving infrastructure and recommendation system engine, enabling real-time item retrieval and scoring using deep tree matching and neural networks. The system distinguishes itself through a multi-task learning framework that optimizes multiple objective functions within a shared representation space. It features a specialized online serving infrastructure that supports dynamic model hot-loading a

Memgraph is an in-memory, distributed graph database designed for high-performance labeled property graph management. It utilizes a Cypher query engine for declarative data retrieval and manipulation, providing a scalable knowledge graph backend that integrates vector search and graph traversals. The system distinguishes itself as a real-time graph analytics platform, employing native C++ and CUDA implementations to execute complex network analysis and dynamic community detection on streaming data. It provides specialized support for AI integration, including GraphRAG capabilities, the constr

ccls is a language server providing code completion, navigation, and semantic analysis for C and C++ projects. It focuses on the indexing and navigation of C-family codebases, enabling users to trace function calls and class hierarchies within a text editor. The project is designed for large scale project indexing, employing parallel source scanning to process massive source trees. It maintains a global symbol index that supports incremental updates as files are saved to disk, ensuring cross-references and symbol lookups remain current. The server covers a range of static code analysis workf

Marksman is a Language Server Protocol implementation for Markdown that provides advanced editor features including autocomplete, go-to-definition, and hover support for markdown files. It serves as an analysis engine to manage documentation and knowledge bases through automated link validation and consistent cross-referencing across multiple documents. The project enables the navigation of complex networks of wiki-style links and headings. It supports content refactoring, allowing users to rename headings and update internal references across multiple files to maintain document integrity dur

This project is a comprehensive framework for constructing, managing, and evaluating knowledge graphs through multi-agent reasoning and deep search capabilities. It provides an end-to-end pipeline that ingests multi-format documents, extracts entities and relationships based on configurable schemas, and maintains structured knowledge bases to support evidence-based retrieval. The system distinguishes itself through its multi-agent orchestration, which decomposes complex queries into parallel research steps and synthesizes long-form reports. It leverages advanced graph-based techniques, includ