jiangxinke/Agentic-RAG-R1
An agentic RAG framework that uses GRPO reinforcement learning to train language models for autonomous retrieval and reasoning tasks.
Not currently ranked — collecting fresh signals.
star history
The project builds a Retrieval-Augmented Generation system enhanced with agentic capabilities through reinforcement learning. It trains a base language model using the GRPO algorithm to perform autonomous document retrieval, multi-step reasoning, and tool use. The framework includes rollout generation for training, inference pipelines for deployment, and evaluation on benchmarks like MedQA to measure medical domain reasoning performance.
Frequently asked
- What is jiangxinke/Agentic-RAG-R1?
- An agentic RAG framework that uses GRPO reinforcement learning to train language models for autonomous retrieval and reasoning tasks.
- Is Agentic-RAG-R1 open source?
- Yes — jiangxinke/Agentic-RAG-R1 is open source, released under the Apache-2.0 license.
- What language is Agentic-RAG-R1 written in?
- jiangxinke/Agentic-RAG-R1 is primarily written in Python.
- How popular is Agentic-RAG-R1?
- jiangxinke/Agentic-RAG-R1 has 422 stars on GitHub.
- Where can I find Agentic-RAG-R1?
- jiangxinke/Agentic-RAG-R1 is on GitHub at https://github.com/jiangxinke/Agentic-RAG-R1.