arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.02423cs.LGcs.AIcs.HCcs.RO
收藏

Uni-RLHF:面向多样化人类反馈的强化学习的通用平台与基准套件

Uni-RLHF: Universal Platform and Benchmark Suite for Reinforcement Learning with Diverse Human Feedback

College of Intelligence and Computing, Tianjin University(天津大学智能与计算学部);Huawei Noah’s Ark Lab(华为诺亚方舟实验室)

Yifu Yuan, Jianye Hao, Yi Ma, Zibin Dong, Hebin Liang, Jinyi Liu, Zhixin Feng, Kai Zhao, Yan Zheng

AI总结:

本文提出Uni-RLHF,一个包含通用多反馈标注平台、大规模众包数据集和模块化基线的综合系统,通过完整工作流程促进基于真实人类反馈的强化学习研究。

摘要:

基于人类反馈的强化学习(RLHF)因通过统一人类偏好来执行任务而无需昂贵的手动奖励设计,受到了广泛关注。在不同环境中考虑多样化的人类反馈类型和各种学习方法至关重要。然而,由于缺乏标准化的标注平台和广泛使用的统一基准,量化具有多样化反馈的RLHF的进展具有挑战性。为了弥补这一差距,我们引入了Uni-RLHF,这是一个为RLHF量身定制的综合系统实现。它旨在提供从真实人类反馈出发的完整工作流程,促进实际问题的开发进展。Uni-RLHF包含三个包:1)通用多反馈标注平台,2)大规模众包反馈数据集,3)模块化离线RLHF基线实现。Uni-RLHF开发了针对各种反馈类型量身定制的用户友好型标注界面,兼容多种主流RL环境。随后,我们建立了系统的众包标注流程,生成了包含30多个热门任务中超过1500万步的大规模标注数据集。通过广泛的实验,收集数据集上的结果展示了与精心设计的手动奖励相比具有竞争力的性能。我们评估了各种设计选择,并提供了关于其优势和潜在改进领域的见解。我们希望构建有价值的开源平台、数据集和基线,以促进基于现实人类反馈的更强大、更可靠的RLHF解决方案的开发。网站可在https://uni-rlhf.github.io/获取。

英文摘要:

Reinforcement Learning with Human Feedback (RLHF) has received significant attention for performing tasks without the need for costly manual reward design by aligning human preferences. It is crucial to consider diverse human feedback types and various learning methods in different environments. However, quantifying progress in RLHF with diverse feedback is challenging due to the lack of standardized annotation platforms and widely used unified benchmarks. To bridge this gap, we introduce Uni-RLHF, a comprehensive system implementation tailored for RLHF. It aims to provide a complete workflow from real human feedback, fostering progress in the development of practical problems. Uni-RLHF contains three packages: 1) a universal multi-feedback annotation platform, 2) large-scale crowdsourced feedback datasets, and 3) modular offline RLHF baseline implementations. Uni-RLHF develops a user-friendly annotation interface tailored to various feedback types, compatible with a wide range of mainstream RL environments. We then establish a systematic pipeline of crowdsourced annotations, resulting in large-scale annotated datasets comprising more than 15 million steps across 30+ popular tasks. Through extensive experiments, the results in the collected datasets demonstrate competitive performance compared to those from well-designed manual rewards. We evaluate various design choices and offer insights into their strengths and potential areas of improvement. We wish to build valuable open-source platforms, datasets, and baselines to facilitate the development of more robust and reliable RLHF solutions based on realistic human feedback. The website is available at https://uni-rlhf.github.io/.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑