ARPO-SFT-54K

Agentic Reinforced Policy Optimization (ARPO) Dataset This repository contains the datasets associated with the paper Agentic Reinforced Policy Optimization (ARPO). ARPO proposes a novel agentic Reinforcement Learning algorithm designed for training multi-turn Large Language Model (LLM)-based agents. It addresses the challenge of balancing intrinsic long-horizon reasoning capabilities with proficiency in multi-turn tool interactions, particularly noting the increased uncertainty in… See the full description on the dataset page: https://huggingface.co/datasets/dongguanting/ARPO-SFT-54K.