Hugging Face releases LeRobot v0.6.0 with new world model policies, reward models, and simulation benchmarks
HIGHLIGHTS
- •Hugging Face release of LeRobot v0.6.0
- •With new world model policies, a new reward models API, and six new simulation benchmarks, including depth sensing, VLM-powered dataset annotation, custom video encoding, cloud training on HF Jobs, and a leaner install.
- •It enables learning to imagine the future and enhances capabilities in simulation and training.
Published:
Official Source (Hugging Face Blog)