Skip to content
repository radar
Agents & Orchestration

🎨ART

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!

⭐ Very popular: 11k stars, gaining about 41 a week

View on GitHub ↗

repo profile

vintage 2025 1 year old
language Python
license Apache-2.0

momentum

total stars 11k
stars added last week +41
HN peak 116 pts 1y 3mo ago
commits / week 5 cooling · -42% vs prior mo
issues closed 50% ~7d to close, median
contributors 27
release cadence monthly
last activity today

durability

backing VC-backed OpenPipe (YC S23)
openness permissive
bus factor 2 concentrated
top-author share 37% 6 mo

bus factor = how many people it takes to cover more than half the commits (6 months). 1 is a solo project; higher means the work is spread across a team. top-author share is the single busiest author's slice of those commits.

since we covered it

monthly average + 305/mo (+4%/mo) · + 3k total since PR#17

why it's a big deal

  • Gives Python developers a way to improve LLM agents on multi-step, real-world tasks through reinforcement learning rather than prompt tweaking alone.
  • RULER auto-generates rewards, so teams can run RL without hand-engineering a reward function for each task.
  • Works with common open models including Qwen, Llama, and GPT-OSS, and integrates with LangGraph and MCP servers for agent workflows.

under the hood

  • Trains with GRPO in a client-server loop where an OpenAI-compatible client runs the agent while a separate GPU server handles inference and training.
  • The four-phase loop runs rollouts, assigns rewards, trains a LoRA on grouped trajectories, then reloads that LoRA into vLLM before resuming; inference is blocked during training.
  • Uses LoRA for parameter-efficient fine-tuning and vLLM for multi-GPU inference, supporting most vLLM and HuggingFace transformers compatible causal models, with Gemma 3 noted as unsupported.

Radar summary, generated from the project's public sources

star history

PR#17 · 7k11k now Mar 2025Aug 2026
  1. PR#17 7k 2025-09-17
  2. now 11k + 3k since first covered

curve is sampled from GitHub's star history, plus our own daily readings since we covered it; 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.

understory

Output and attention are roughly in balance.

-2 understory score output 59 · clout 61
Aug 2025 Jul 2026
  • output, commits & releases
  • clout, star velocity

output = commits & releases; clout = star velocity, both 0 to 100 monthly indices; the gap where output runs above clout is the understory. The understory →

covered in

  • PR#17 2025-09-17 below the radar

    Automated Reasoning Trainer

similar projects

compare these →
  • Unsloth

    6.8× the stars

    Local UI to run and train LLMs and diffusion models, including Qwen3.8, Kimi K3, MiniMax-H3, Gemma 4, DeepSeek-V4, FLUX and more.

    72k ACTIVE
  • 🧠 MaxKB

    2.1× the stars · GPL-3.0 licensed

    🔥 MaxKB is an open-source platform for building enterprise-grade agents.

    23k ACTIVE
  • 🛠️ Agent Lightning

    The absolute trainer to light up AI agents.

    17k ACTIVE

comments

Sign in with GitHub to add your blip on ART.

loading comments…