roller-es3
Roller sprint 2m, round 3 of evolution-strategies fine-tuning on the Arena roller simulator (CPU MuJoCo, BAM servos), continued from es2. Base: Pollen reference roller policy fine-tuned with PPO (warm1). Race-only: training seeds 101-400, held-out 1001-1030 = 1.669 s, board seeds 1-5 never searched. Simulation only, no hardware test.
A perpetual policy for the microduck (61-D observation, 14 actions, 50 Hz, on rollers). Runs until told otherwise — a gait, loaded into a policy slot.
Run it on a robot
sudo robotctl policy load <slot> stefo0/microduck-roller-es3
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:
main - commit:
423289dc8
- Downloads last month
- 17