Openvla Is An Open-source Generalist Robotics Model

OpenVLA is an open-source vision-language-action (VLA) model that has been trained on a diverse collection of real-world robot demonstrations. It outperforms other similar models on robotics tasks and can easily be fine-tuned for generalization in multi-task environments. OpenVLA makes foundation models more accessible and customizable for a broader range of companies and research labs.