jump
Jump: crouch, airborne, land standing. Trained 2026-11-10 on an M4 MacBook Air (custom task Mjlab-Jump-Flat-MicroDuck, 300 rounds x 256 envs). Probe: 100% of 192 episodes airborne, 0% fallen endings.
A episodic policy for the microduck (61-D observation, 14 actions, 50 Hz). Runs 4.0 s and returns itself to a standing pose.
Run it on a robot
sudo robotctl policy add jump chair510/microduck-jump
robotctl robot do jump
The observation normalizer is baked into policy.onnx; feed raw observations.
manifest.json follows schema 2 of the microduck policy manifest (docs/policy-manifest.md in the daemon repo).
Training
- repo:
pollen-robotics/microduck_rl - branch:
develop - commit:
0daab160d - exported from a checkout with uncommitted changes
- Downloads last month
- 19