Awesome GitHub RepositoriesDynamic Data Source Connectors

JimuReport is an open-source reporting and dashboard engine designed to be embedded directly into Spring Boot applications. Its core identity centers on generating data reports and full-screen dashboards from natural language descriptions, eliminating the need for manual design. The platform also provides a conversational query interface that translates plain-language questions into database queries, returning results as tables and charts without requiring SQL knowledge. What distinguishes JimuReport is its integration of AI skills that can be installed with a single command, enabling report

Hotwire Turbo is a server-driven navigation and HTML streaming framework that intercepts link clicks and form submissions to fetch pages in the background and replace content without full browser reloads. It provides a Turbo Frame component that scopes page regions into independent contexts, enabling partial page updates where only a specific area of the page navigates or loads content. The framework includes a page cache and morph system that stores recently visited pages for instant restoration and applies minimal DOM changes on refresh, preserving scroll position and element state. The fra

SpringBlade is a development framework and platform designed for building multi-tenant SaaS applications. It provides a comprehensive scaffold for both Spring Cloud microservices and monolithic Spring Boot architectures, enabling the rapid construction of enterprise-grade software. The platform distinguishes itself through integrated LLM orchestration and industrial IoT management. It features an LLM orchestration platform that combines large language models with knowledge bases and visual AI agent workflows, alongside an IoT hub for device connectivity, state synchronization, and edge flow o

Openblocks is a low-code platform for building custom internal tools. It provides a visual interface where users can assemble applications by dragging and dropping pre-built components onto a canvas, connecting them to databases and APIs without writing code. The platform distinguishes itself through its architecture for embedding and reuse. Entire application pages can be rendered as native React components inside other applications, replacing traditional iframe-based embedding. Custom components and queries can be bundled into reusable modules for use across multiple applications, and an au

Ignite ist ein High-Level-Trainings-Framework für PyTorch-Neuronale Netze, das als Trainings-Engine und Deep-Learning-Lifecycle-Manager dient. Es bietet ein strukturiertes System zur Organisation und Automatisierung von Trainings- und Evaluierungsschleifen, verwaltet Daten-Iteratoren und löst Event-Handler bei bestimmten Meilensteinen während des Modelltrainings aus. Das Projekt zeichnet sich durch eine umfassende Suite von Tools für verteiltes Training und Modellevaluierung aus. Es enthält Dienstprogramme zur Synchronisierung von Gradienten und zur Koordination kollektiver Kommunikation über mehrere GPUs oder Knoten hinweg sowie eine Evaluierungssuite zur Berechnung von Leistungsmetriken und zur Durchführung von k-facher Kreuzvalidierung. Die breiteren Funktionen umfassen die Automatisierung von Trainings-Workflows, einschließlich Learning-Rate-Scheduling, Early Stopping und Hyperparameter-Optimierung. Das Framework bietet zudem Observability-Tools für Experiment-Tracking, Profiling der Ausführungszeit und Mixed-Precision-Training zur Optimierung der Speicherauslastung. Mechanismen zur Statuspersistenz sind enthalten, um Modell-Checkpoints zu verwalten und Trainingssitzungen wiederherzustellen. Containerisierte Umgebungen sind verfügbar, um die Bereitstellung und Einrichtung zu vereinfachen.

Arroyo is a high-performance stream processing platform built in Rust. It executes continuous SQL queries on streaming data with event-time semantics, enabling accurate windowed aggregations, joins, and stateful computations on unbounded event streams. The platform uses native Rust execution for high throughput and low latency, with periodic checkpointing for exactly-once fault tolerance and horizontal scaling across distributed workers. The system integrates deeply with Kafka for reading and writing topics with exactly-once delivery and supports change data capture (CDC) from MySQL and Postg