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 Qwen3.5, GPT-OSS, Llama, and more!

65
/ 100
Established

9,149 stars. Actively maintained with 20 commits in the last 30 days.

No Package No Dependents
Maintenance 20 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 19 / 25

How are scores calculated?

Stars

9,149

Forks

789

Language

Python

License

Apache-2.0

Last pushed

Apr 10, 2026

Commits (30d)

20

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/fine-tuning/OpenPipe/ART"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.