AgentRL
Multitask, multiturn RL for LLM agents; Ray-based scaling, rollout/actor workers—for teams that want to train agents, not just run them.
trainingpython
- Stars
- 302
- Adoption surface
- complex
- Autonomy
- headless
- Recovery
- resumable
- License
- ✅ open-source
- Category
- Multi-agent and orchestration
Repository ↗ Example: Async GRPO trainer ↗