OpenPipe/ART
Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen3.6, GPT-OSS, Llama, and more!
Build with Backblaze B2
SDKs, agent skills, IDE extensions, and reference pipelines from Backblaze Labs. All open source.
Loading star history...
Where it stands
Rank, growth pace and release cadence relative to everything GStars tracks.
5 releases in the last 90 days, the latest 1 day ago, with 26 contributors on GitHub.
Use Cases & Benefits
- Train multi-step AI agents for real-world tasks using reinforcement learning with the GRPO algorithm.
- Enables automatic agent behavior improvement and tool usage optimization without manual reward engineering.
- Use for building adaptive ReAct-style agents that improve through experience in complex multi-step workflows.
- Use for integrating reinforcement learning training into existing Python applications with minimal code changes.
- Use for training language models like Qwen, GPT-OSS, and Llama to perform domain-specific tasks such as email retrieval or game playing.
Should you depend on it?
License terms, project hygiene and risks, from GitHub metadata and GStars' snapshots.
License
Apache 2.0
Permissive: commercial use, modification and redistribution are allowed with attribution; includes an explicit patent grant from contributors.
Community files
- README: present
- Contributing guide: present
Watch-outs
No watch-outs from GStars' checks.
General guidance from repository metadata, not legal advice. How we score
From the README
Train multi-step agents for real-world tasks using GRPO.
ART's LangGraph integration enables you to train sophisticated ReAct-style agents that improve through reinforcement learning. Build agents that reason, use tools, and adapt their behavior over time without manual prompt engineering.
ART is an open-source RL framework that improves agent reliability by allowing LLMs to learn from experience. ART provides an ergonomic harness for integrating GRPO into any python application. For a quick hands-on introduction, run one of the notebooks below. When you're ready to learn more, check out the docs.
Daily snapshots
Stars gained per day, and how forks and open issues moved, from GStars' daily snapshots
Forks
1.0K+42
Open issues
186+65
Compared with peers in AI Agents
Fastest-growing repositories GStars tracks in this cohort for the week of Nov 8, by 7-day star growth
| Repository | Stars | +7d | Health | Last release |
|---|---|---|---|---|
| morluto/rea | 450.0K | +22.9K | 75 | 16 days ago |
| vercel-labs/agent-desktop | 21.4K | +8.7K | 75 | 5 days ago |
| milind-soni/OpenMausBot | 20.6K | +8.4K | 75 | today |
| stablyai/orca | 121.4K | +6.7K | 75 | today |
| thedotmack/claude-mem | 115.8K | +4.7K | 75 | today |
| OpenPipe/ART | 10.8K | — | 50 | 1 day ago |
| Category median | 115.8K | +8.4K | 75 | — |
Discover Repositories
Search across tracked repositories by name or description