DAPO: An Open-Source LLM Reinforcement Learning System at Scale
By: Qiying Yu , Zheng Zhang , Ruofei Zhu and more
Potential Business Impact:
Teaches AI to solve hard math problems better.
Inference scaling empowers LLMs with unprecedented reasoning ability, with reinforcement learning as the core technique to elicit complex reasoning. However, key technical details of state-of-the-art reasoning LLMs are concealed (such as in OpenAI o1 blog and DeepSeek R1 technical report), thus the community still struggles to reproduce their RL training results. We propose the $\textbf{D}$ecoupled Clip and $\textbf{D}$ynamic s$\textbf{A}$mpling $\textbf{P}$olicy $\textbf{O}$ptimization ($\textbf{DAPO}$) algorithm, and fully open-source a state-of-the-art large-scale RL system that achieves 50 points on AIME 2024 using Qwen2.5-32B base model. Unlike previous works that withhold training details, we introduce four key techniques of our algorithm that make large-scale LLM RL a success. In addition, we open-source our training code, which is built on the verl framework, along with a carefully curated and processed dataset. These components of our open-source system enhance reproducibility and support future research in large-scale LLM RL.
Similar Papers
Training Long-Context, Multi-Turn Software Engineering Agents with Reinforcement Learning
Machine Learning (CS)
Makes AI agents solve complex coding problems.
A New DAPO Algorithm for Stock Trading
Computational Engineering, Finance, and Science
Makes trading computers smarter and faster.
Comparative Analysis and Parametric Tuning of PPO, GRPO, and DAPO for LLM Reasoning Enhancement
Artificial Intelligence
Teaches computers to think better and solve problems.