Skip to content
RobothecaBeta
Log in

UniLab

  • Software
  • Humanoids and robot learning
  • Framework

UniLab is a configurable framework for robot reinforcement learning that lets tasks be defined once in Hydra and run across several physics simulators and hardware backends through one CLI.

Made by Motphys.

Source on GitHub: github.com/Motphys/UniLab (opens in a new tab). Facts read at commit c8e1d73 (opens in a new tab).

Licenses

Code
Apache-2.0 (opens in a new tab)

Read by Robotheca from LICENSE (opens in a new tab) (Apache-2.0) at commit c8e1d73 (opens in a new tab), on 22 Oct 2026.

Commercial use
Yes

What it does

  • Tasks are assembled from manager terms in Hydra YAML, so variants built from existing terms often need no new Python environment class.
  • A single task contract lets the chosen simulator be swapped with a flag while training and evaluation stay the same.
  • Simulation can run on CPU, native, or external workers while the learner runs on a separate accelerator.
  • Supports SAC and FlashSAC, with simulation data collection overlapping learner updates for replay-based off-policy training.
  • Includes a train and eval CLI, sim-to-sim deployment across several simulators, and demo policies such as g1 walking and motion tracking.
  • Documented paths cover Windows, Apple Silicon macOS, Linux CUDA, AMD ROCm and Intel XPU, with a support matrix grading backend maturity.

Written by Robotheca from the project's README (opens in a new tab) at commit c8e1d73.

Getting started

Difficulty
Advanced

Maintenance

Status
Active: a commit in the last 90 days
Newest commit
22 Oct 2026 on main
Latest release
unilab 1.5.0 (22 Oct 2026)

Checked on GitHub on 22 Oct 2026.

Links

Robots that can run it

No robot on Robotheca is linked to it yet. Links come from a robot's robot.yaml, from its dependency files read as text, or from Robotheca's editors.

Robotheca lists software as a reference: nothing is hosted or run here, and every fact links to where it was read.