ROBOTO_ORIGIN · published weights
walk (default velocity-tracking policy)
Record
- Robot
- ROBOTO_ORIGIN roboparty/roboto-origin
- Framework
- Isaac Lab + RSL-RL (roboparty_train), ONNX via roboparty_inference · Isaac Lab
- Policy license
- not stated by the package · robot software repos: GPL-3.0-only
- File
- robots/rpo/models/policy.onnx
- Note
- Started with `./tools/start_robot.sh --robot rpo --policy default`
Not in the tab yet
To replay this policy in the browser we need an ONNX export of the named checkpoint, the observation and action contract copied from the authors' own sim2sim or deploy script, and a MuJoCo model of the robot that the contract matches. Then it must walk 60 s in our runtime without a fall. Until then the source link is the way in.
Other ROBOTO_ORIGIN policies · 6
- walk, AMP styleIsaac Lab + RSL-RL AMP
- walk, attention encoderIsaac Lab + RSL-RL
- motion tracking: wave / dance (BeyondMimic)Isaac Lab BeyondMimic
- get up from the groundIsaac Lab BeyondMimic-style motion tracking
- interruptible walkingIsaac Lab + RSL-RL
- parkour with depthIsaac Lab + RSL-RL