Today, I found that Hugging Face now supports Agents. I decided to give it a try!
An Agent is a system that uses an LLM as its engine and has the ability to access external tools.
So, the system has two core components: the engine (LLM) and the tools.
It’s quite easy to implement an Agentic RAG.
Using the ReactJsonAgent mode from `transformers.agents`, here’s a code block showing the execution log.
