craigm26/microduck-duckbatch-b002-128x128
A walking policy for Pollen Robotics' Microduck:
26,254 parameters (61-128-128-14), distilled from Pollen's
default walker (velstand.onnx, 197,774 parameters) by
duckbatch, batch b002-student-size-longer, attempt a02.
Drive it in your browser (Pollen's simulator loads it from this repo).
Simulation only. It has never run on a real robot. A student that matches the teacher in mjlab is a candidate for a hardware test, not a result about hardware.
Measured
mjlab Mjlab-VelStand-Flat-MicroDuck with domain randomization, stumble pushes and observation noise on;
the task's deliberate prone spawns and topple pushes are off for the walking numbers and on
(every episode prone) for the get-up numbers.
30 s per seed on held-out seeds
2001, 2002, 2003; mean ± spread over seeds. The teacher ran
in the same eval.
| This policy | Teacher | |
|---|---|---|
| Falls per minute | 0.46 ± 0.06 | 0.14 ± 0.02 |
| Time fallen | 1.12% ± 0.34% | 0.43% ± 0.10% |
| Planar velocity error (m/s) | 0.182 ± 0.001 | 0.174 ± 0.004 |
| Yaw rate error (rad/s) | 0.407 ± 0.004 | 0.368 ± 0.003 |
| Gets up from prone within 6 s | 97.1% ± 0.5% | 96.9% ± 1.5% |
| Time to get up (s) | 0.56 ± 0.02 | 0.56 ± 0.01 |
| Parameters | 26,254 | 197,774 |
| FLOPs per step | 51,968 | 393,728 |
| 1-thread latency, p50 (x86 laptop) | 10.8 us | 29.0 us |
Use
- Browser:
https://pollen-robotics-microduck-simulator.hf.space/?move=craigm26/microduck-duckbatch-b002-128x128 - Robot:
robotctl policy load walk craigm26/microduck-duckbatch-b002-128x128(manifest schema 2, walk slot) - Contract:
obs[1,61] -> actions[1,14], the normalizer baked in, the same as every Pollen policy.
Full record (training trace, every judge decision, the decision-model answers):
https://github.com/craigm26/duckbatch, records/b002-student-size-longer/.