robocasa

🎁 A collection of utilities for LeRobot.

An efficient inference runtime for Cosmos 3 Nano / Edge robot policies

A simple, efficient, and extensible evaluation framework for robot policies across diverse benchmarks, supporting both local and remote execution

Run and compare robot VLAs and world-action models in LIBERO and RoboCasa with reproducible local evidence.

Example usage of a set of simulation benchmarks

Paired perturbation and recovery datasets for robot learning, with native simulator evidence, replay and audit tools.

Confidence-aware world-aligned 3D scene memory and SmolVLA integration for RoboCasa

Action-conditioned latent world model and experiment code for RoboCasa365.

MOPS Data Generation

AMD ROCm Physical AI training, evaluation, simulation, rendering, and reproducible evidence suite