ART - Python项目推荐

Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen2.5, Qwen3, Llama, and more!

项目统计：⭐ 8,232 stars | 🍴 659 forks📈 今日新增 52 stars

编程语言：Python

标签：agent, agentic-ai, grpo, llms, lora

开源协议：Apache-2.0

GitHub 链接：https://github.com/OpenPipe/ART 项目主页：https://art.openpipe.ai