| title | FastOPD: On-Policy Distillation for Lightweight VLA Deployment |
|---|---|
| authors | Yoojin Oh, Jeongsol Kim, Yeonwoo Seo, Jangho Park, Seonghyun Jin, Sunwoo Park, Youngmin Kim, Youngjun Jun, Kyumin Choi, Jong Chul Ye |
| date | 2026-10-02 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2610.02832 |
| affiliation | KAIST AI |
| upvotes | 2 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| primer | Vision-language-action models |
| abstract | Vision-Language-Action (VLA) foundation models have scaled rapidly to enhance manipulation performance and generalizability, but this scaling incurs high computational costs that render real-world deployment increasingly challenging. Existing approaches typically mitigate this issue by designing smaller architectures or reducing the iterative denoising steps in flow-based policies. In this work, we propose FastOPD, a foundation-to-lightweight VLA framework that enables the practical deployment of large-scale VLAs through efficient on-policy distillation. Specifically, FastOPD adapts a flow map for single-state teacher supervision and combines it with a self-consistency objective to construct a compact student that learns the teacher dynamics. Furthermore, we theoretically demonstrate that minimizing this objective allows the distilled student to recover a distribution on par with that induced by an ideal few-step teacher model. We evaluate FastOPD across diverse foundation policies in simulation and real-world experiments. On LIBERO, FastOPD retains 84% of the performance of π_{0.5} with only two inference steps, reducing inference latency by 78.1% while outperforming existing few-step distillation baselines in average success rate. With LingBot-VLA as the teacher, FastOPD improves the single-step success rate over the base student by 15.9 percentage points on RoboTwin 2.0. We further demonstrate its applicability to a World Action Model (WAM) and deploy a compact student distilled from MolmoAct2 on a real robot. |
cite
@misc{oh2026fastopd,
title={FastOPD: On-Policy Distillation for Lightweight VLA Deployment},
author={Yoojin Oh and Jeongsol Kim and Yeonwoo Seo and Jangho Park and Seonghyun Jin and Sunwoo Park and Youngmin Kim and Youngjun Jun and Kyumin Choi and Jong Chul Ye},
year={2026},
eprint={2610.02832},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2610.02832}
}