CO-RFT: Efficient Fine-Tuning of Vision-Language-Action Models through Chunked Offline Reinforcement Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Huang, Dongchi, Fang, Zhirui, Zhang, Tianle, Li, Yihang, Zhao, Lin, Xia, Chunhe
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913974405038080
author Huang, Dongchi
Fang, Zhirui
Zhang, Tianle
Li, Yihang
Zhao, Lin
Xia, Chunhe
author_facet Huang, Dongchi
Fang, Zhirui
Zhang, Tianle
Li, Yihang
Zhao, Lin
Xia, Chunhe
contents Vision-Language-Action (VLA) models demonstrate significant potential for developing generalized policies in real-world robotic control. This progress inspires researchers to explore fine-tuning these models with Reinforcement Learning (RL). However, fine-tuning VLA models with RL still faces challenges related to sample efficiency, compatibility with action chunking, and training stability. To address these challenges, we explore the fine-tuning of VLA models through offline reinforcement learning incorporating action chunking. In this work, we propose Chunked RL, a novel reinforcement learning framework specifically designed for VLA models. Within this framework, we extend temporal difference (TD) learning to incorporate action chunking, a prominent characteristic of VLA models. Building upon this framework, we propose CO-RFT, an algorithm aimed at fine-tuning VLA models using a limited set of demonstrations (30 to 60 samples). Specifically, we first conduct imitation learning (IL) with full parameter fine-tuning to initialize both the backbone and the policy. Subsequently, we implement offline RL with action chunking to optimize the pretrained policy. Our empirical results in real-world environments demonstrate that CO-RFT outperforms previous supervised methods, achieving a 57% improvement in success rate and a 22.3% reduction in cycle time. Moreover, our method exhibits robust positional generalization capabilities, attaining a success rate of 44.3% in previously unseen positions.
format Preprint
id arxiv_https___arxiv_org_abs_2508_02219
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CO-RFT: Efficient Fine-Tuning of Vision-Language-Action Models through Chunked Offline Reinforcement Learning
Huang, Dongchi
Fang, Zhirui
Zhang, Tianle
Li, Yihang
Zhao, Lin
Xia, Chunhe
Robotics
Machine Learning
Vision-Language-Action (VLA) models demonstrate significant potential for developing generalized policies in real-world robotic control. This progress inspires researchers to explore fine-tuning these models with Reinforcement Learning (RL). However, fine-tuning VLA models with RL still faces challenges related to sample efficiency, compatibility with action chunking, and training stability. To address these challenges, we explore the fine-tuning of VLA models through offline reinforcement learning incorporating action chunking. In this work, we propose Chunked RL, a novel reinforcement learning framework specifically designed for VLA models. Within this framework, we extend temporal difference (TD) learning to incorporate action chunking, a prominent characteristic of VLA models. Building upon this framework, we propose CO-RFT, an algorithm aimed at fine-tuning VLA models using a limited set of demonstrations (30 to 60 samples). Specifically, we first conduct imitation learning (IL) with full parameter fine-tuning to initialize both the backbone and the policy. Subsequently, we implement offline RL with action chunking to optimize the pretrained policy. Our empirical results in real-world environments demonstrate that CO-RFT outperforms previous supervised methods, achieving a 57% improvement in success rate and a 22.3% reduction in cycle time. Moreover, our method exhibits robust positional generalization capabilities, attaining a success rate of 44.3% in previously unseen positions.
title CO-RFT: Efficient Fine-Tuning of Vision-Language-Action Models through Chunked Offline Reinforcement Learning
topic Robotics
Machine Learning
url https://arxiv.org/abs/2508.02219