qhjqhj00/MemoRAG
A RAG framework using memory-based knowledge discovery to enhance retrieval and response generation with long-context LLMs.
Not currently ranked — collecting fresh signals.
star history
MemoRAG is an RAG framework that leverages a memory model to achieve global understanding of entire databases. It recalls query-specific clues from memory to enhance evidence retrieval and generate more accurate, contextually rich responses. The system includes training scripts, a Lite mode for processing millions of tokens, and was published at theWebConf 2025.
Frequently asked
- What is qhjqhj00/MemoRAG?
- A RAG framework using memory-based knowledge discovery to enhance retrieval and response generation with long-context LLMs.
- Is MemoRAG open source?
- Yes — qhjqhj00/MemoRAG is open source, released under the Apache-2.0 license.
- What language is MemoRAG written in?
- qhjqhj00/MemoRAG is primarily written in Python.
- How popular is MemoRAG?
- qhjqhj00/MemoRAG has 2.3k stars on GitHub.
- Where can I find MemoRAG?
- qhjqhj00/MemoRAG is on GitHub at https://github.com/qhjqhj00/MemoRAG.