roller-sprint-2m-roller-2-e
Another checkpoint of my second roller run. It leaves the skating behind and runs for speed, on a recipe shared by TheSauceman. It finishes every run in about 1.5 seconds.
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> GinDuck/microduck-roller-sprint-2m-roller-2-e
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
- task_id:
Mjlab-RollerSprint2m-MicroDuck - repo:
https://github.com/pollen-robotics/microduck-challenges - branch:
race-2 - commit:
b44f7ffc9 - checkpoint:
5000 - seed:
1 - base:
mjlab-microduck 0.1.0 @ 1efd853be - started:
2026-11-02T13:27:08Z
Reproduce
Same code, same uv.lock, same command, same seed. Training it again yields a comparable policy, not the same weights: GPU reinforcement learning is not bit-reproducible across machines.
git clone https://github.com/pollen-robotics/microduck-challenges
cd microduck-challenges
git checkout b44f7ffc9
uv sync
uv run train Mjlab-RollerSprint2m-MicroDuck --env.scene.num-envs 2048 --agent.max-iterations 6000 --agent.seed 1 --agent.logger tensorboard --agent.run-name roller-2
- Downloads last month
- 10