Skip to content
Repository Radar

ART

BTR ACTIVE STEADY

OpenPipe/ART · homepage ↗

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!

  • llms
  • lora
  • reinforcement-learning
  • agent
  • agentic-ai
  • grpo
stars 10k
last activity 3d ago
open issues 62
language Python
license Apache-2.0
latest release v0.5.17
momentum · per month since covered + 314/mo (+4%/mo) · + 3k total since PR#17

metrics as of today

star history

Exact curve on star-history.com ↗
PR#17 · 7k10k★ now Mar 2025Jul 2026
  1. PR#17 7k★ 2025-09-17
  2. now 10k★ + 3k since first covered

curve is sampled from GitHub's star history; the dashed stretch is before we first covered it, the solid line since. figures at coverage are the numbers we printed then (approx.), current count is live.

covered in

  • PR#17 2025-09-17 below the radar

    Automated Reasoning Trainer

// comments

COMING SOON

Sign in with GitHub to weigh in on ART. We're wiring this up; check back soon.