LIBERO-Pro2.0 provides robot manipulation demonstrations for imitation learning and vision-language-action research. It contains 40,470 stored demonstrations across 32 base tasks in four LIBERO suites, including base tasks and perturbation variants. Each demonstration includes a language instruction, scene and wrist camera observations, robot states, and actions. Trajectories are generated by a privileged-state oracle and distributed in native HDF5 format. - 8,258,504 steps, with two 128 × 128 RGB camera views and 7-dimensional actions. - 6,745 review videos: five successful examples per task/variation instance. - Training data only; no validation or test split is provided. Repeated…
Independent publisher
Zhou Xueyang
zhouxueyang
Models in Library0
Datasets in Library1
Models on Hugging Face—
Followers5
Compare the published evidence
This profile connects the 1 model and dataset record shown below to their upstream cards and sources. The library counts describe SAVRN's coverage, rather than this publisher's entire output.
Published tasks: Robotics.
Reported licenses: cc-by-nc-4.0. Open a record to review its specific license and conditions.
Compare model specifications, source dates, and hardware estimates on the individual records. Publisher identity verification is separate from a model's quality, deployment readiness, or suitability for your work.