Welcome to OpenMLRL
We are a group of researchers focusing on optimizing multi-LLM systems with reinforcement learning. We develop open-source frameworks and tools to enrich this community.
Our missions are:
- Training LLM agents to cooperate effectively to solve tasks together
- Improving the performance and scalability of RL methods in multi-LLM systems
- Enabling multi-turn interactions between agents and other models to strengthen reasoning
Open-Source Library for Cooperative Multi-LLM Reinforcement Learning