AI Skill Hub 强烈推荐：本地脑知识库 是一款优质的AI工具。AI 综合评分 8.0 分，在同类工具中表现稳健。如果你正在寻找可靠的AI工具解决方案，这是一个值得深入了解的选择。
本地脑知识库 是一款基于 Python 开发的开源工具，专注于 chromadb、cuda、gpu-acceleration 等核心功能。作为 GitHub 开源项目，它拥有活跃的社区支持和持续的版本迭代，代码完全透明可审计，支持本地部署以保护数据隐私。无论是个人使用还是集成到企业工作流，都能提供稳定可靠的解决方案。
本地脑知识库 是一款基于 Python 开发的开源工具，专注于 chromadb、cuda、gpu-acceleration 等核心功能。作为 GitHub 开源项目，它拥有活跃的社区支持和持续的版本迭代，代码完全透明可审计，支持本地部署以保护数据隐私。无论是个人使用还是集成到企业工作流，都能提供稳定可靠的解决方案。
# 方式一：pip 安装（推荐）
pip install local-brain-retrieval-augmented-generation
# 方式二：虚拟环境安装（推荐生产环境）
python -m venv .venv
source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install local-brain-retrieval-augmented-generation
# 方式三：从源码安装（获取最新功能）
git clone https://github.com/rrg1225/Local-Brain-Retrieval-Augmented-Generation
cd Local-Brain-Retrieval-Augmented-Generation
pip install -e .
# 验证安装
python -c "import local_brain_retrieval_augmented_generation; print('安装成功')"
# 命令行使用
local-brain-retrieval-augmented-generation --help
# 基本用法
local-brain-retrieval-augmented-generation input_file -o output_file
# Python 代码中调用
import local_brain_retrieval_augmented_generation
# 示例
result = local_brain_retrieval_augmented_generation.process("input")
print(result)
# local-brain-retrieval-augmented-generation 配置文件示例（config.yml） app: name: "local-brain-retrieval-augmented-generation" debug: false log_level: "INFO" # 运行时指定配置文件 local-brain-retrieval-augmented-generation --config config.yml # 或通过环境变量配置 export LOCAL_BRAIN_RETRIEVAL_AUGMENTED_GENERATION_API_KEY="your-key" export LOCAL_BRAIN_RETRIEVAL_AUGMENTED_GENERATION_OUTPUT_DIR="./output"
<p align="center"> <img src="https://img.shields.io/badge/Python-3.10+-3776AB?style=for-the-badge&logo=python&logoColor=white" alt="Python"> <img src="https://img.shields.io/badge/PyTorch-2.6-EE4C2C?style=for-the-badge&logo=pytorch&logoColor=white" alt="PyTorch"> <img src="https://img.shields.io/badge/CUDA-12.8-76B900?style=for-the-badge&logo=nvidia&logoColor=white" alt="CUDA"> <img src="https://img.shields.io/badge/GPU-RTX_5060-ED1C24?style=for-the-badge&logo=nvidia&logoColor=white" alt="RTX 5060"> <img src="https://img.shields.io/badge/License-MIT-yellow?style=for-the-badge" alt="License"> <img src="https://img.shields.io/badge/Privacy-100%25_Local-2EA44F?style=for-the-badge&logo=lock&logoColor=white" alt="Privacy"> </p>
<p align="center"> <b>100% Local Inference · Zero Data Leakage · Built for Developers</b><br> An industrial-grade Retrieval-Augmented Generation (RAG) system that turns your<br> entire codebase, technical docs, and architecture diagrams into a conversational <i>second brain</i>. </p>
<p align="center"> <sub>English · <a href="#中文">中文</a></sub> </p>
---
🔒 Privacy IsolationAll embeddings and reranking run on your local GPU. Code snippets, documents, and conversations never leave your machine. ChromaDB multi-collection architecture provides physical isolation per project space. |
⚡ Hybrid RetrievalBM25 (code-aware tokenizer) + Vector (BGE) + RRF fusion. The custom |
🚀 Hardware-OptimizedTuned for NVIDIA RTX 5060 · PyTorch 2.6 · CUDA 12.8. Concurrent batch ingestion with thread-pool workers, incremental mtime-based scanning, and BM25 instance caching that avoids O(N) vocabulary rebuilds on every chat turn. |
📄 Multi-Format ParsingPyMuPDF for PDF, python-docx for Word, Gemini Vision for architecture diagrams/screenshots, and tree-sitter |
🛡️ Graceful DegradationLLM quota exhausted? The system auto-fails over from Gemini to Qwen (and back) mid-conversation without losing context. BM25 init failure? Falls back to pure vector retrieval. Resilient by design. |
🪝 Live File WatchingWatchdog monitors workspace directories with configurable debounce. Files are incrementally re-indexed on change; deleted files are pruned from the vector store. Ghost node cleanup runs on every startup. |
---
pip install -r requirements.txt ```
```bash
pip install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu128
conda create -n bendirag python=3.11 -y conda activate bendirag
```bash
高性能的离线知识库，优先考虑隐私和安全
该工具未明确声明开源协议，商业使用前请联系原作者确认授权范围，避免侵权风险。
AI Skill Hub 为第三方内容聚合平台，本页面信息基于公开数据整理，不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后，再部署至生产环境，并做好必要的安全评估。
总体来看，本地脑知识库 是一款质量优秀的AI工具，在同类工具中具备一定竞争力。AI Skill Hub 将持续追踪其更新动态，建议收藏备用，结合自身场景选择合适时机引入使用。
| 原始名称 | Local-Brain-Retrieval-Augmented-Generation |
| Topics | chromadbcudagpu-accelerationpython |
| GitHub | https://github.com/rrg1225/Local-Brain-Retrieval-Augmented-Generation |
| 语言 | Python |
收录时间：2026-06-04 · 更新时间：2026-06-04 · License：未公布 · AI Skill Hub 不对第三方内容的准确性作法律背书。