microduck-rl running
Training repository with 39 registered tasks and a published straight-running policy holding 1.844 m/s over 10 seconds and 1.806 m/s over 30 under a heading-hold controller, stating plainly that the 2.0 m/s target was not met. Sim-only.
https://github.com/zhoumiaosen/microduck-rl61 OBSERVATIONS 14 ACTIONS 50 HZ COMPATIBILITY NOT INDEPENDENTLY VERIFIED
AWESOME MICRODUCK INDEX
microduck-rl running
Training repository with 39 registered tasks and a published straight-running policy holding 1.844 m/s over 10 seconds and 1.806 m/s over 30 under a heading-hold controller, stating plainly that the 2.0 m/s target was not met. Sim-only.
Project Overview
Train, evaluate, and export reinforcement-learning policies for Pollen Robotics' MicroDuck, a small biped with 14 actuators. Built with mjlab, MuJoCo Warp, and PPO , with 39 registered tasks covering locomotion and additional behaviors.
Download the trained model · Watch the running video · Setup · Train · Validation
The current release reaches 1.844 m/s straight progress over 10 seconds and 1.806 m/s over 30 seconds in simulation, using a heading-hold controller. The sustained 2.0 m/s target was not met .
Task
| Field | Value |
|---|---|
| Source | GITHUB SOURCE |
| Runtime contract | 61 observations / 14 actions / 50 Hz |
| Verification | Pending independent Duckio review |
| Artifact | Not indexed |
Source Coverage
The upstream README includes: Published running policy, Setup, Try the published model, Train and resume, Policy interface, Earlier running experiment, Development, Attribution and license.
Intended Use
Use this entry to assess the project and locate its canonical implementation. Read the complete README, inspect the manifest and artifacts, and reproduce the source's evaluation before assigning simulation or hardware verification.