A curated data mixture recipe of approximately 25 carefully weighted datasets from Open X-Embodiment.
No structured details available.
No tags available.
A collection of key datasets used for training embodied AI models, including robot manipulation and vision-language-action models.