搜索结果：RAG

RAG-Anything is an open-source framework developed by HKU researchers that enables unified multimodal retrieval-augmented generation, allowing AI systems to understand and retrieve knowledge from text

RAG-Anything is a comprehensive all-in-one multimodal document processing RAG system that seamlessly handles text, images, tables, equations, and other content types within a unified framework, elimin

本文提出了一种超越传统文档检索的RAG框架，将其演化为动态优化引擎。该框架通过动态检索和组合多种上下文（如定制指令、少样本示例），并基于交互反馈实时优化检索策略，优先选择高效用内容。系统形成闭环学习，能自动纠正错误、强化成功输出，使LLM无需重新训练即可持续改进，实现从静态检索到实时自适应优化的转变。

This research systematically evaluates quantization and dimensionality reduction techniques for RAG vector embeddings, finding that float8 quantization combined with PCA offers optimal 8x compression

Retrieval-Augmented Generation (RAG) is an AI framework that enhances Large Language Models (LLMs) by integrating real-time information retrieval from external knowledge bases, addressing limitations

Satya is an offline-first educational platform that integrates Retrieval-Augmented Generation (RAG) with the Phi 1.5 language model, designed to run locally on standard hardware (4GB RAM) without internet dependency, specifically addressing educational infrastructure gaps in rural areas. 原文翻译： Satya是一个离线优先的教育平台，集成了检索增强生成（RAG）与Phi 1.5语言模型，旨在无需互联网依赖，在标准硬件（4GB RAM）上本地运行，专门解决农村地区的教育基础设施差距。

The GPT-RAG Solution Accelerator provides enterprise-ready architecture templates and deployment assets for building secure, scalable Retrieval-Augmented Generation (RAG) solutions powered by AI Agents on Azure, incorporating Zero-Trust security, Responsible AI, and end-to-end observability. 原文翻译： GPT-RAG解决方案加速器为企业提供即用型架构模板和部署资产，用于在Azure平台上构建安全、可扩展的检索增强生成（RAG）解决方案，由AI代理驱动，融合零信任安全、负责任AI和端到端可观测性。

Microsoft Research's GraphRAG enhances LLM capabilities by generating knowledge graphs from private datasets, significantly improving question-answering performance and enabling whole-dataset reasoning through graph machine learning. 原文翻译： 微软研究院的GraphRAG通过从私有数据集中生成知识图谱，显著提升大型语言模型的能力，通过图机器学习大幅改善问答性能并实现全数据集推理。

GraphRAG (Graphs + Retrieval Augmented Generation) is a technique for richly understanding text datasets by combining text extraction, network analysis, and LLM prompting and summarization into a single end-to-end system. 原文翻译： GraphRAG（图+检索增强生成）是一种通过将文本提取、网络分析、LLM提示和摘要整合到单一端到端系统中，来深入理解文本数据集的技术。

LLM Knowledge Graph Builder transforms unstructured data into Neo4j knowledge graphs, offering superior accuracy over traditional RAG through GraphRAG's relationship-aware retrieval mechanism. 原文翻译： LLM知识图谱构建器将非结构化数据转换为Neo4j知识图谱，通过GraphRAG的关系感知检索机制，提供比传统RAG更优越的准确性。