Skip to main content

A memory management system for large language models

Project description

LLM Memory

一个专为大语言模型设计的记忆管理系统,支持基于重要性的记忆筛选和持久化存储。

安装

使用以下命令安装:
pip install llm-memory

使用示例:

from llm_memory import CombinedMemory

# 初始化记忆系统
memory = CombinedMemory(
    max_size=100,                # 最大记忆容量
    system_prompt="你是一个助手",  # 系统提示词
    uid="user123",              # 用户ID,用于数据库存储
    recent_size=20              # 保留最近对话轮数
)

# 添加消息
memory.add_message("user", "你好", importance=1)
memory.add_message("assistant", "你好!有什么我可以帮你的吗?", importance=1)

# 获取上下文
context = memory.get_context()

# 保存到数据库
memory.save_memory_to_db()

参数说明:

  • system_prompt: 系统提示词,可选
  • uid: 用户唯一标识,用于数据库存储,可选
  • max_size: 最大记忆容量,必需
  • recent_size: 保留最近对话轮数,默认为20轮

特性:

  • 基于重要性的记忆管理
  • 自动保持最近的对话(可配置保留轮数)
  • 支持数据库持久化
  • 可配置的系统提示词

数据持久化:

当提供 uid 参数时,会自动将对话历史保存到本地数据库(使用 TinyDB)。数据库文件默认保存为 memory.json。

依赖:

Python >= 3.7 tinydb >= 4.7.0

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

llm_memory-0.2.1.tar.gz (4.3 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

llm_memory-0.2.1-py3-none-any.whl (4.2 kB view details)

Uploaded Python 3

File details

Details for the file llm_memory-0.2.1.tar.gz.

File metadata

  • Download URL: llm_memory-0.2.1.tar.gz
  • Upload date:
  • Size: 4.3 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.1.0 CPython/3.9.6

File hashes

Hashes for llm_memory-0.2.1.tar.gz
Algorithm Hash digest
SHA256 35826dfe875b2822ad9afed17884b4a38b3995426dac95454dd08bef4c69a214
MD5 dc06a721ae4459e1e4bad2983bd7efa0
BLAKE2b-256 cb61ef761ad4bbec5328663c804bdec99c9ebb53fabbcfe4df5b9c463a4c0fcb

See more details on using hashes here.

File details

Details for the file llm_memory-0.2.1-py3-none-any.whl.

File metadata

  • Download URL: llm_memory-0.2.1-py3-none-any.whl
  • Upload date:
  • Size: 4.2 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.1.0 CPython/3.9.6

File hashes

Hashes for llm_memory-0.2.1-py3-none-any.whl
Algorithm Hash digest
SHA256 b9ef8b9255922306273c6caa5d7f045b9fdec72ce21927e1e49e7c660af9173d
MD5 76189e31e8c185b836c07ef4c3cce08b
BLAKE2b-256 7bee83a072aa3db6e8e51f75c6369c63a124dee6a57cb462d9b7f8131cbcd9f8

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page