Welcome to the homepage of RAG4j/p. RAG is short for Retrieval Augmented Generation. RAG has become famous for its power to help Large Language Models (LLMs) generate text that is more coherent and relevant to the given context. The 4j has become famous for being a Java framework. Therefore, we adopted the 4p as well, the Python version.

In the world of Python, many frameworks are available that can help you use RAG. But in the land of Java, there is not much choice. We like the frameworks Langchain and Langchain4j a lot, and we got some inspiration from them. So why another framework?

The main focus of RAG4j/p is teaching people the power of RAG. We want to make it easy to use RAG and understand how RAG works. We use this project for workshops. With all the necessary code, learning about the different components becomes much more accessible. Playing around, adjusting, and extending them with new versions of components is easy to learn. With the knowledge from this project, you can easily switch to other, more extensive frameworks that give you much more than RAG. We stick to RAG.

One part of the project is something not all frameworks have: a way to determine the quality of your RAG system. We got our inspiration from the TruLens project. We tried to keep it as simple as possible to use.

Where to from here? The code is available on Github RAG4j. If you are more into Python, we’ve got you covered; there is Github RAG4p

We hope you like the project. Feel free to contact us if you have questions or remarks.

Posts

subscribe via RSS