Skip to content
@intellistream

IntelliStream

Research group focused on stream processing, AI systems, and intelligent databases.

IntelliStream Research Group

Total Stars

Website GitHub Organization

专注于流处理、AI系统与智能数据库的研究与开发

Focused on Stream Processing, AI Systems, and Intelligent Databases


🌟 SAGE 项目生态系统 | SAGE Project Ecosystem

SAGE (Streaming-Augmented Generative Execution) 是 IntelliStream 面向 AI 数据处理与推理流水线的主框架。 当前核心代码已经重新收敛到一个主仓库中,围绕统一的 L1→L5 分层架构组织开发、测试与发布。

SAGE is IntelliStream's primary framework for AI data processing and inference pipelines, organized as a single main repository with a unified L1→L5 layered architecture.

SAGE GitHub Stars PyPI

  • Monorepo packages: sage, sage-common, sage-platform, sage-kernel, sage-libs, sage-middleware, sage-cli, sage-tools
  • In-tree benchmarking: benchmark configs, experiment scripts, and evaluation assets now live alongside the main repo
  • Public docs: maintained in SAGE-Pub

🏗️ SAGE 分层架构(L1→L5)| SAGE Layered Architecture

L1  sage-common
 ↓
L2  sage-platform
 ↓
L3  sage-kernel     sage-libs
 ↓
L4  sage-middleware
 ↓
L5  sage-cli        sage-tools

核心主仓库 | Core Monorepo

层级 / Layer 组件 / Components 作用 / Role
L1 sage-common 配置、日志、协议、公共工具
L2 sage-platform 队列、存储、服务与运行基础设施抽象
L3 sage-kernel, sage-libs 流式运行时、调度、算法与数据处理能力
L4 sage-middleware 向量库、记忆系统、网络算子与运行时绑定组件
L5 sage-cli, sage-tools CLI、开发工作流、质量检查与维护工具

附加内容:

  • benchmark/:主仓库内维护的评测配置、实验脚本和结果资产
  • packages/sage/:聚合安装入口与统一发布包

配套仓库 | Companion Repositories

这些仓库仍然独立维护,但围绕主 SAGE 仓库协同工作:

仓库 / Repo 用途 / Purpose
SAGE-Pub 公共文档、网站与发布文档源
wheelwright 通用 Python 包发布工具
sage-studio 可视化工作流与交互式体验入口
sage-edge 边缘部署与聚合网关

说明:公共 profile 仅保留对外可见且仍建议独立使用的配套仓库;内部团队仓库不在此处展示。


🧠 vLLM-HUST 生态 | vLLM-HUST Ecosystem

vLLM-HUST 是 IntelliStream 后续重点推进的 LLM 推理研发主线,基于 vLLM 进行增强与工程化扩展。

vLLM-HUST is IntelliStream's primary LLM inference development line, built on top of vLLM with engineering-focused extensions.

vllm-hust (core)
  ├─ vllm-hust-workstation (product/runtime workspace)
  └─ vllm-hust-website (public site and docs portal)

� 研究评测仓库 | Research Benchmark Repos

每个评测仓库独立维护,专注于评估对应子系统的性能与正确性。

Each benchmark repo is independently maintained and focuses on evaluating the corresponding subsystem.

仓库 / Repo 评测对象 / Evaluates 说明 / Description
sage-agentic-tooluse-benchmark Stars sage-agentic-tooluse Agent 工具调用选择算法评测
sage-rag-benchmark Stars sage-rag RAG 检索与重排管道评测
sage-refiner-benchmark Stars sageRefiner 上下文压缩与精炼算法评测
sage-memory-benchmark Stars neuromem 记忆管理与检索性能评测
SAGE-DB-Bench Stars sage-anns / sageVDB 流式 ANN 向量数据库评测
LibAMM Stars sage-amms 近似矩阵乘法算法评测

�🚀 其他研究项目 | Other Research Projects

流处理系统 | Stream Processing Systems

  • MorphStream ⭐ 141 - [ICDE'20, SIGMOD'23, TKDE'24] 可扩展的事务性流处理引擎 | Scalable transactional stream processing engine
  • AllianceDB ⭐ 16 - [SIGMOD'21] 并行数据库系统 | Parallel database system

基准测试与工具 | Benchmarks & Tools

  • Sesame ⭐ 26 - [SIGMOD'23] 数据流聚类实证研究 | Data stream clustering empirical study
  • PDSC - 并行数据流聚类基准 | Parallel data stream clustering benchmark

机器学习与AI | Machine Learning & AI

  • SentiStream ⭐ 7 - [EMENLP'23] 情感分析流处理 | Sentiment analysis stream processing
  • StreamLearning - 流式学习框架 | Stream learning framework

资源与文档 | Resources & Documentation


📖 快速开始 | Quick Start

安装 SAGE | Install SAGE

# PyPI 安装 | Install from PyPI
pip install isage

# 开发安装 | Development installation
git clone https://github.com/intellistream/SAGE.git
cd SAGE
./quickstart.sh --dev --yes

简单示例 | Simple Example

from sage.kernel.api.local_environment import LocalEnvironment
from sage.libs.io.source import FileSource
from sage.middleware.operators.rag import DenseRetriever, QAPromptor, OpenAIGenerator
from sage.libs.io.sink import TerminalSink

# 创建执行环境 | Create execution environment
env = LocalEnvironment("rag_pipeline")

# 构建声明式管道 | Build declarative pipeline
(
    env.from_source(FileSource, {"file_path": "questions.txt"})
    .map(DenseRetriever, {"model": "sentence-transformers/all-MiniLM-L6-v2"})
    .map(QAPromptor, {"template": "Answer based on: {context}\nQ: {query}\nA:"})
    .map(OpenAIGenerator, {"model": "gpt-3.5-turbo"})
    .sink(TerminalSink)
)

# 执行管道 | Execute pipeline
env.submit()

详细文档请访问:SAGE Documentation

For detailed documentation, visit: SAGE Documentation


🤝 参与贡献 | Contributing

我们欢迎各种形式的贡献!请查看各个仓库的 CONTRIBUTING.md 文件了解详情。

We welcome contributions of all kinds! Please check the CONTRIBUTING.md file in each repository for details.


📞 联系我们 | Contact Us


📄 许可证 | License

各项目许可证详见各仓库的 LICENSE 文件。大多数项目采用 MIT 或 Apache 2.0 许可证。

License details can be found in each repository's LICENSE file. Most projects use MIT or Apache 2.0 licenses.


⭐ 如果我们的项目对您有帮助,请给我们一个 Star!

If our projects help you, please give us a Star!

Popular repositories Loading

  1. MorphStream MorphStream Public

    This project aims at building a scalable transactional stream processing engine on modern hardware. It allows ACID transactions to be run directly on streaming data. It shares similar project visio…

    C 140 8

  2. StreamProcessing_ReadingList StreamProcessing_ReadingList Public

    stream processing reading list

    69 10

  3. SAGE SAGE Public

    Shell 39 9

  4. Sesame Sesame Public

    [SIGMOD'23] Data Stream Clustering: An In-depth Empirical Study [ICDM'24] MOStream: A Modular and Self-Optimizing Data Stream Clustering Algorithm

    C++ 26 6

  5. CANDOR-Benchmark CANDOR-Benchmark Public

    C++ 25 9

  6. AllianceDB AllianceDB Public

    C++ 16 8

Repositories

Showing 10 of 74 repositories

Top languages

Loading…

Most used topics

Loading…