6.2 KiB
PromethAI-Memory
Open-source framework that manages memory for AI Agents and LLM apps
Share promethAI Repository
The Motivation
Browsing the database of theresanaiforthat.com, we can observe around 7000 new, mostly semi-finished projects in the field of applied AI, whose development is fueled by new improvements in foundation models and open-source community contributions.
It seems it has never been easier to create a startup, build an app, and go to market… and fail.
AI apps currently being pushed out still mostly feel and perform like demos.
To address this issue, dlthub and prometh.ai will collaborate on productionizing a common use-case, progressing step by step. We will utilize the LLMs, frameworks, and services, refining the code until we attain a clearer understanding of what a modern LLM architecture stack might entail.
We go on a journey and propose a new way to reason about Language Architecture for Agents based on cognitive sciences.
Read more on our blog post prometh.ai
Or check this Princeton paper released after our demo
PromethAI-Memory Repo Structure
The repository contains a set of folders that represent the steps in the evolution of the modern data stack from POC to production
Level 1 - CMD script to process PDFs
We introduce the following concepts:
- Structured output with Pydantic
- CMD script to process custom PDFs
Level 2 - Memory Manager implemented in Python
We introduce the following concepts:
- Long Term Memory
- Short Term Memory
- Episodic Buffer
- Attention Modulators
The code at this level contains:
- Simple PDF ingestion
- FastAPI
- Docker Image
- Memory manager
- Langchain-based Agent Simulator
- Data schema
How to use
Each of the folders contains a README to get started.