OpenPipe/ART
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!
Builder
OpenPipe
OpenPipe • individual
Stars
9,081
Using upstream star count
Forks
775
Using upstream fork count
Open Issues
0
Activity Score
0/100
14 commits in 30d
Created
Mar 10, 2025
Project creation date
README Summary
Agent Reinforcement Trainer (ART) is a framework for training multi-step AI agents on real-world tasks using Group Relative Policy Optimization (GRPO). It provides on-the-job training capabilities for agents to improve their performance through reinforcement learning. The system supports multiple language models including Qwen3.5, GPT-OSS, and Llama.
AI Dev Skills
Unmapped
Tags
Taxonomy
Deployment Context
Modalities
Skill Areas
Recent Activity
Updated 27 days ago
7 Days
0
30 Days
14
90 Days
123
Quality
prototype- Quality
- medium
- Maturity
- prototype
Categories
PM Skills
Languages
Timeline
- Project created
- Mar 10, 2025
- Forked
- Mar 12, 2026
- Your last push
- 27 days ago
- Upstream last push
- 6 days ago
- Tracked since
- Mar 17, 2026
Similar Repos
pgvector cosine similarity · $0
Loading…