---
tags:
- microduck
- robotics
- reinforcement-learning
- onnx
library_name: microduck
pipeline_tag: robotics
base_model: pollen-robotics/microduck-roller-sprint-2m-reference
base_model_relation: finetune
---

# roller-es2

2 m Roller Sprint skater: Pollen's roller reference fine-tuned by evolution strategies on the Arena's own roller scorer (x86-verified).

A **perpetual** policy for the [microduck](https://github.com/pollen-robotics/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

```bash
sudo robotctl policy load <slot> juicenv/microduck-roller-es2
```

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).

A remix of [pollen-robotics/microduck-roller-sprint-2m-reference](https://huggingface.co/pollen-robotics/microduck-roller-sprint-2m-reference).

## Training

- **repo**: `pollen-robotics/microduck_rl`
- **branch**: `lean`
- **commit**: `5a3636302`