Revise README for clarity and accuracy
Some checks failed
Mark stale issues and pull requests / stale (push) Has been cancelled
Some checks failed
Mark stale issues and pull requests / stale (push) Has been cancelled
Updated project description and core features section in README.md.
This commit is contained in:
parent
6c35fd732a
commit
fed559dbfc
16
README.md
16
README.md
@ -1,6 +1,5 @@
|
||||
|
||||
<div align="center">
|
||||
<h1>语析 - 基于大模型的知识库与知识图谱智能体开发平台</h1>
|
||||
<h1>语析 - 结合知识库与知识图谱的多租户 Harness 平台</h1>
|
||||
|
||||
[](https://github.com/xerrors/Yuxi/blob/main/docker-compose.yml)
|
||||
[](https://github.com/xerrors/Yuxi/issues)
|
||||
@ -24,18 +23,10 @@
|
||||
## 核心特性
|
||||
|
||||
- **智能体开发**:基于 LangGraph,支持子智能体、Skills、MCPs、Tools 与中间件机制
|
||||
- **知识库(RAG)**:多格式文档上传,支持 Embedding / Rerank 配置及知识库评估
|
||||
- **知识库(RAG)**:多格式文档解析,支持 Embedding / Rerank 配置及知识库评估
|
||||
- **知识图谱**:基于 LightRAG 的图谱构建与可视化,支持属性图谱并参与智能体推理
|
||||
- **平台与工程化**:Vue + FastAPI 架构,支持暗黑模式、Docker 与生产级部署
|
||||
|
||||
|
||||
## 你可以用语析做什么?
|
||||
|
||||
- 构建 **面向真实业务的 RAG + 知识图谱智能体**
|
||||
- 将 PDF / Word / Markdown / 图片快速转化为可推理的知识库
|
||||
- 自动(LightRAG)或手动构建知识图谱,并用于智能体推理
|
||||
- 使用 LangGraph v1 构建多智能体 / 子智能体系统
|
||||
|
||||
## 最新动态
|
||||
|
||||
|
||||
@ -201,10 +192,11 @@ docker compose up --build
|
||||
本项目参考并引用了以下优秀开源项目,在此致以诚挚的感谢:
|
||||
|
||||
- [LightRAG](https://github.com/HKUDS/LightRAG) - 直接引入作为图谱构建与检索的基础包
|
||||
- [DeepAgents](https://github.com/IDEA-CCNL/DeepAgents) - 直接引入作为深度智能体框架
|
||||
- [DeepAgents](https://github.com/langchain-ai/deepagents) - 直接引入作为深度智能体框架
|
||||
- [DeerFlow](https://github.com/bytedance/deer-flow) - 参考了其 Sandbox 智能体架构的实现思路
|
||||
- [RAGflow](https://github.com/infiniflow/ragflow) - 参考了其文档 Text Chunking 的分块策略
|
||||
- [LangGraph](https://github.com/langchain-ai/langgraph) - 多智能体编排框架,本项目的核心架构基础
|
||||
- [QwenPaw](https://github.com/agentscope-ai/QwenPaw) - 参考模型配置与个人文件区域设计
|
||||
|
||||
## 参与贡献
|
||||
|
||||
|
||||
Loading…
Reference in New Issue
Block a user