FastOPD: On-Policy Distillation for Lightweight VLA Deployment
FastOPD: On-Policy Distillation for Lightweight VLA Deployment
FastOPD 提出一套以高效策略內蒸餾將大型 VLA 模型部署到輕量化模型的框架。在 LIBERO 上,模型只需兩個推理步驟,便保留 π_{0.5} 84% 的效能,並將推理延遲降低 78.1%。以 LingBot-VLA 為教師模型時,FastOPD 在 RoboTwin 2.0 上的單步成功率較基礎學生模型高 15.9 個百分點。
Published on Oct 2
Authors:
,
,
,
,
,
,
,
,
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.
View arXiv page View PDF Project page Add to collection
Models citing this paper 0
No model linking this paper
Cite arxiv.org/abs/2610.02832 in a model README.md to link it from this page.
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2610.02832 in a dataset README.md to link it from this page.
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2610.02832 in a Space README.md to link it from this page.
Collections including this paper 0
No Collection including this paper
Add this paper to a collection to link it from this page.
來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co