skills.rest
nebius
nebiusOFFICIAL
Explore all skills by this author →
WebsiteGitHub

sim-to-real

Orchestrate robotics simulation, policy training, and evaluation workflows via declarative YAML.

17|8|Updated Apr 7, 2026
Search Tags:#workflow#orchestration#simulation#robotics#machine-learning#physical-ai
One-click install
npx skills add https://github.com/nebius/nebius-physical-ai --skill sim-to-real-nebius
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sim-to-real
Source: https://github.com/nebius/nebius-physical-ai/tree/main/skills/workflows/sim-to-real
Command: npx skills add https://github.com/nebius/nebius-physical-ai --skill sim-to-real-nebius

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill addresses the complexity of managing robotics data pipelines by providing a unified, declarative framework for simulation, policy training, and evaluation.

Core Features & Use Cases

  • Workflow Orchestration: Manages the lifecycle of data from simulation to real-world deployment using standardized run IDs.
  • Tool Integration: Seamlessly connects with workbench tools like Genesis, LeRobot, and VLM-eval for end-to-end experimentation.
  • Use Case: A robotics team can use this skill to automate the training of a new navigation policy, ensuring that synthetic data generation, model training, and performance evaluation are executed consistently across cloud infrastructure.

Quick Start

Execute the sim-to-real workflow by triggering the pipeline with your specific run configuration and target S3 prefix.

Frequently Asked Questions about sim-to-real

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What is the best way to manage robotics data pipelines from simulation to deployment?▼

You can manage robotics data pipelines by adopting a unified, declarative framework that connects simulation, synthetic data generation, policy training, and evaluation. This standardizes run IDs to automate experimentation and ensure consistent lifecycle management.

Can I automate synthetic data generation and policy training using declarative YAML specifications?▼

Yes, workflow orchestration seamlessly integrates with workbench-compatible tools like Genesis, LeRobot, and VLM-eval. This tool integration enables end-to-end experimentation by connecting synthetic data generation directly with model training and evaluation.

Does workflow orchestration for physical AI work with tools like Genesis and LeRobot?▼

Yes, you can automate synthetic data generation and policy training by defining declarative YAML workflow specifications. Triggering the pipeline with your specific run configuration and target S3 prefix executes the entire sim-to-real workflow consistently.

How do I orchestrate end-to-end physical AI workflows for robotics?▼

You need the Nebius Physical AI workbench environment to ensure experiment reproducibility and consistent artifact management. This environment provides the necessary infrastructure orchestration and workbench-compatible tool integration for robotics data flows.

Related Skills

Explore More Software Engineering Skills→
nebiusnebius
17

submit-workflow

Validate and submit SkyPilot workflows for Nebius Physical AI clusters.

Official
Advanced
EurecaMomentEurecaMoment

benchclaw-stage5-eval

Automate model evaluation and report generation for BenchClaw benchmarks.

Official
Advanced
nebiusnebius
17

train-policy

Coordinates robot-policy training lifecycle across LeRobot, Isaac Lab, and SONIC via YAML specifications and Nebius GPU/S3 infrastructure.

Official
Advanced
zli5460zli5460
2

weights-and-biases

Logs ML metrics, hyperparameters, outputs, and manages models via Werkzeug and Bias tools.

Community
Intermediate