🚀 Quick Start
Environment setup, quick checks, evaluation, and training workflows.
StarVLA Documentation
Build, train, and evaluate Vision-Language-Action models with a modular, extensible codebase.
What is StarVLA?
Section titled “What is StarVLA?”StarVLA is a modular and flexible codebase for developing Vision-Language Models (VLMs) into Vision-Language-Action (VLA) models. Each component (model, data, trainer, configuration, evaluation) is designed for high cohesion and low coupling, enabling plug-and-play research and fast iteration.
Docs at a Glance
Section titled “Docs at a Glance”📖 Project Overview
What StarVLA is, current capabilities, and key links.
🧩 Lego-like Design
The modular design principles behind StarVLA.
🤖 Model Zoo
Released models and finetuning checkpoints.
📚 FAQ
Common questions about configs, backbones, and training.