Fine-tune NVIDIA's DreamZero (14B World Action Model) on your own robot dataset with one command, on Amazon SageMaker - View it on GitHub