Dexbotic: Open-Source Vision-Language-Action Toolbox

TL;DR

Dexbotic is an open-source, versatile toolbox for Vision-Language-Action research, supporting multiple policies, easy experiment development, and enhanced pretrained models to advance embodied intelligence studies.

Contribution

It introduces a comprehensive, open-source VLA toolbox with multi-policy support, easy experiment customization, and improved pretrained models for the research community.

Findings

Supports multiple VLA policies simultaneously

Enables quick development of new VLA experiments

Provides stronger pretrained models for better performance

Abstract

In this paper, we present Dexbotic, an open-source Vision-Language-Action (VLA) model toolbox based on PyTorch. It aims to provide a one-stop VLA research service for professionals in the field of embodied intelligence. It offers a codebase that supports multiple mainstream VLA policies simultaneously, allowing users to reproduce various VLA methods with just a single environment setup. The toolbox is experiment-centric, where the users can quickly develop new VLA experiments by simply modifying the Exp script. Moreover, we provide much stronger pretrained models to achieve great performance improvements for state-of-the-art VLA policies. Dexbotic will continuously update to include more of the latest pre-trained foundation models and cutting-edge VLA models in the industry.