Manage robot policy training, evaluation, and inference with LeRobot.
npx skills add https://github.com/nebius/nebius-physical-ai --skill lerobot
Please help me install this Agent Skill. Skill: lerobot Source: https://github.com/nebius/nebius-physical-ai/tree/main/skills/tools/lerobot Command: npx skills add https://github.com/nebius/nebius-physical-ai --skill lerobot
This Skill addresses the complexity of managing robot policy training, evaluation, and inference workflows by providing a unified interface for the LeRobot framework.
Use the lerobot skill to deploy the default training environment for your robot policy workflow.
High-intent search queries and answers about installing and using this skill.
Robot policy training and inference workflows are managed through a unified interface that handles the end-to-end lifecycle, including model evaluation and serving using the LeRobot framework.
Yes, the LeRobot framework supports diverse policy architectures, including ACT, Diffusion Policy, and SmolVLA, allowing you to train and serve various models within Nebius-based infrastructure.
You can convert simulation outputs to Hugging Face format using the dataset conversion capabilities built into the LeRobot policy lifecycle management workflow.
Yes, you can seamlessly switch between specific LeRobot versions, such as 0.5.1 and 0.6.0, to ensure environment parity and facilitate version-specific image deployment.
To deploy a robot policy for real-time inference, you can use the serving interfaces provided to evaluate the trained model against a golden dataset and serve it within the Nebius infrastructure.
You need access to Nebius-based infrastructure to effectively train robot policies like Diffusion Policy on a cluster, evaluate them, and serve the resulting models.
Andrew-GirgisTrain and fine-tune language and vision models on Hugging Face Jobs.
KAIROS7-team3Implement SROS2 DDS security, network segmentation, and host hardening for robotic systems.
pmcdowallManage Hugging Face Hub resources via hf CLI operations.
ActiveInferenceInstituteExtracts GNN features and trains/evaluates ML models across multiple frameworks.
