Summary
LeRobot v0.5.0 has been released, described by its developers as the largest update to the open-source robotics library so far. The headline feature is support for the Unitree G1 humanoid robot, marking LeRobot's first whole-body control system for humanoids. The release adds six new policies: Pi0-FAST (an autoregressive VLA model), Real-Time Chunking, Wall-X, X-VLA, SARM, and PEFT. Performance improvements include streaming video encoding and up to 10x faster image training. A new EnvHub component allows simulation environments to be loaded directly from the Hugging Face Hub. Infrastructure updates include Python 3.12+, Transformers v5 compatibility, and integration with NVIDIA IsaacLab-Arena. The project is open source and available on GitHub at huggingface/lerobot. This update strengthens LeRobot's position as a leading library for embodied AI and robot learning research.
LeRobot v0.5.0 has been released, and it is the largest update to the project so far.
Core Highlights
- Unitree G1 humanoid robot: LeRobot's first humanoid robot system with whole-body control
- 6 new policies: Pi0-FAST (autoregressive VLA), Real-Time Chunking, Wall-X, X-VLA, SARM, and PEFT
- Performance improvements: streaming video encoding and 10x faster image training
- EnvHub: load simulation environments directly from the Hugging Face Hub
- Infrastructure: Python 3.12+, Transformers v5, and NVIDIA IsaacLab-Arena
Project: https://github.com/huggingface/lerobot
This page is an English static mirror generated for search and AI citation.
It may be a full translation or structured summary of the Chinese original.
Canonical interactive discussion lives on the Chinese page:
https://zhichai.net/topic/177168842