一个基于 Rust 的超高性能认知图谱记忆引擎,支持毫秒级激活扩散与大规模联想检索。
Project description
Pero-Memory-Core (pero-rust-core)
🚀 一个为 AI 记忆而生的极致性能引擎。
基于 Rust 开发,专为本地化、低功耗、大规模 AI Agent 记忆检索场景设计。
🌟 核心特性
- 千万级联想秒开:基于 Spreading Activation(激活扩散)算法,在 2000 万节点、2000 万关联边的图谱上,5 步联想检索仅需 0.18ms。
- 深度记忆回溯:支持高达 50 步的深度联想,在 1 秒内可处理数十万个关联节点的能量传播。
- 内存极致优化:使用 Rust CSR 变体结构,内存占用仅为同类向量数据库的 1/10。
- 安全性防护:内置正则表达式清洗器,原生防止 ReDoS 攻击,并支持意图向量的动态安全验证。
- 开箱即用:提供纯净的 Python 接口,无需配置复杂的数据库环境。
📊 性能表现
| 任务 | 规模 | 耗时 |
|---|---|---|
| 图谱构建 | 4000 万边 | ~24.0s |
| 浅层联想 (5步) | 4000 万规模 | 0.18ms |
| 深度联想 (50步) | 4000 万规模 | 830ms |
| 文本清洗 | 10万字符 | 0.6ms |
🛠 安装方法 (即将发布)
pip install pero-memory-core
💻 快速开始
from pero_rust_core import CognitiveGraphEngine
# 初始化引擎
engine = CognitiveGraphEngine()
# 批量添加记忆关联 (src_id, target_id, strength)
engine.batch_add_connections([(1, 2, 0.8), (2, 3, 0.5)])
# 执行记忆扩散联想
# initial_scores: 初始激活点, steps: 扩散步数
results = engine.propagate_activation({1: 1.0}, steps=5)
print(f"联想到的相关记忆点: {results}")
📜 许可证
MIT License
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
pero_memory_core-0.1.0.tar.gz
(13.9 kB
view details)
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file pero_memory_core-0.1.0.tar.gz.
File metadata
- Download URL: pero_memory_core-0.1.0.tar.gz
- Upload date:
- Size: 13.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: maturin/1.11.2
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
41ace6cc451f4e4e5435c70197b578c12b3c550e4f7c4206b8bffb92a5f2682f
|
|
| MD5 |
29dbe507d8318253b248e4bee744a096
|
|
| BLAKE2b-256 |
250c4cb5d9cc4276d6ee2c71402e690058626fbd66631c00a62dbaf0090a3679
|
File details
Details for the file pero_memory_core-0.1.0-cp310-cp310-win_amd64.whl.
File metadata
- Download URL: pero_memory_core-0.1.0-cp310-cp310-win_amd64.whl
- Upload date:
- Size: 903.2 kB
- Tags: CPython 3.10, Windows x86-64
- Uploaded using Trusted Publishing? No
- Uploaded via: maturin/1.11.2
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
32d2e8f19990d66c053052bccf9bd93e5003e0260119756b13e6fece417e3732
|
|
| MD5 |
e18affbc947d3e10a17f6efd3e29bd77
|
|
| BLAKE2b-256 |
211166de3a7b09287999dd135f4d41a4740bbededcb81e234042dbb53e193653
|