Teleoperated bimanual manipulation demonstrations collected in MuJoCo with a BimanualYAM robot, packaged in LeRobot v3.0 format for pi0.5 training. Three RGB views per frame — camerafront, cameraleft, cameraright (224x224, H.264). Robot state and action are both 14-dimensional: 6 arm joints plus 1 gripper per arm. A 128-dimensional non-robot state carries object/scene information for diagnostics. Actions are stored as absolute joint targets. The training recipe converts the 12 arm dimensions to deltas relative to the current action-chunk anchor while the 2 gripper dimensions stay absolute. The normalization statistics in meta/stats.json are computed in that 50-step relative space — do not…
Independent publisher
MuchenLiu
mliu48
Models in Library0
Datasets in Library1
Models on Hugging Face—
Followers—
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: apache-2.0. 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.