Skip to main content
The lerobot-info command displays comprehensive system information useful for debugging LeRobot issues and reporting bugs.

​
Command

lerobot-info
Location: src/lerobot/scripts/lerobot_info.py

​
Overview

This utility:
  • Shows LeRobot and dependency versions
  • Reports system platform and Python version
  • Displays GPU and CUDA information
  • Lists installed LeRobot scripts
  • Runs without requiring LeRobot dependencies
  • Outputs in markdown format for easy copying

​
Output Example

lerobot-info
Example output: 
- LeRobot version: 2.1.0
- Platform: Linux-6.5.0-35-generic-x86_64-with-glibc2.35
- Python version: 3.11.9
- Huggingface Hub version: 0.24.0
- Datasets version: 2.19.0
- Numpy version: 1.26.4
- FFmpeg version: 6.1.1
- PyTorch version: 2.3.0+cu121
- Is PyTorch built with CUDA support?: True
- Cuda version: 12.1
- GPU model: NVIDIA GeForce RTX 4090
- Using GPU in script?: <fill in>
- lerobot scripts: ['lerobot-calibrate', 'lerobot-dataset-viz', 'lerobot-edit-dataset', 'lerobot-eval', 'lerobot-find-cameras', 'lerobot-find-joint-limits', 'lerobot-find-port', 'lerobot-imgtransform-viz', 'lerobot-info', 'lerobot-record', 'lerobot-replay', 'lerobot-setup-can', 'lerobot-setup-motors', 'lerobot-teleoperate', 'lerobot-train', 'lerobot-train-tokenizer']

​
Information Displayed

​
LeRobot Information

​
LeRobot version
str
Installed LeRobot package version or “N/A” if not installed.
​
lerobot scripts
list
All installed command-line scripts (e.g., lerobot-train, lerobot-record).

​
System Information

​
Platform
str
Operating system and architecture (e.g., Linux-6.5.0, Windows-10, macOS-14.0).
​
Python version
str
Python interpreter version.

​
Dependency Versions

​
Huggingface Hub version
str
Version of huggingface_hub package.
​
Datasets version
str
Version of datasets package.
​
Numpy version
str
Version of numpy package.
​
FFmpeg version
str
FFmpeg command-line tool version (required for video encoding).

​
PyTorch and GPU Information

​
PyTorch version
str
PyTorch version including CUDA/ROCm build info.
​
Is PyTorch built with CUDA support?
bool
Whether PyTorch has CUDA support compiled in.
​
Cuda version
str
CUDA version PyTorch was built with.
​
GPU model
str
GPU device name (e.g., “NVIDIA GeForce RTX 4090”).
​
Using GPU in script?
str
Placeholder for user to fill in when reporting issues.

​
Use Cases

​
Bug Reports

When reporting issues on GitHub:
  1. Run lerobot-info
  2. Copy the output
  3. Paste into your issue description
  4. Fill in “Using GPU in script?” field
Example: 
## Environment

```text
- LeRobot version: 2.1.0
- Platform: Linux-6.5.0-35-generic-x86_64-with-glibc2.35
- Python version: 3.11.9
...
- Using GPU in script?: Yes, using CUDA device 0

​
Debugging Installation Issues

Check if all dependencies are installed: 
lerobot-info
Look for “N/A” values:
  • LeRobot version: N/A → LeRobot not installed
  • PyTorch version: N/A → PyTorch not installed
  • FFmpeg version: N/A → FFmpeg not installed
  • Cuda version: N/A → CUDA not available

​
Verifying GPU Setup

Check CUDA availability: 
lerobot-info | grep -E "CUDA|GPU"
Example output: 
- Is PyTorch built with CUDA support?: True
- Cuda version: 12.1
- GPU model: NVIDIA GeForce RTX 4090
If CUDA is not available:
  • “Is PyTorch built with CUDA support?: False” → Install CUDA-enabled PyTorch
  • “Cuda version: N/A” → Install CUDA toolkit
  • “GPU model: N/A” → No GPU detected or driver issue

​
Comparing Environments

When collaborating, compare system info: 
# On working system
lerobot-info > working_system.txt

# On problematic system  
lerobot-info > broken_system.txt

# Compare
diff working_system.txt broken_system.txt

​
Checking Script Availability

Verify all LeRobot scripts are installed: 
lerobot-info | grep "lerobot scripts"
Expected scripts:
  • lerobot-calibrate
  • lerobot-dataset-viz
  • lerobot-edit-dataset
  • lerobot-eval
  • lerobot-find-cameras
  • lerobot-find-joint-limits
  • lerobot-find-port
  • lerobot-imgtransform-viz
  • lerobot-info
  • lerobot-record
  • lerobot-replay
  • lerobot-setup-can
  • lerobot-setup-motors
  • lerobot-teleoperate
  • lerobot-train
  • lerobot-train-tokenizer

​
Troubleshooting

​
FFmpeg Not Found

Output: 
- FFmpeg version: N/A
Solution: 
# Ubuntu/Debian
sudo apt install ffmpeg

# macOS
brew install ffmpeg

# Conda
conda install -c conda-forge ffmpeg

# Verify
ffmpeg -version

​
PyTorch Not Using GPU

Output: 
- Is PyTorch built with CUDA support?: False
Solution: Install CUDA-enabled PyTorch: 
# For CUDA 12.1
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121

# For CUDA 11.8
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

# Verify
python -c "import torch; print(torch.cuda.is_available())"

​
Version Parsing Failed

Output: 
- PyTorch version: Installed (version not found)
Cause: Package is installed but version information is unavailable. Impact: Usually not a problem; package is functional.

​
Programmatic Usage

from lerobot.scripts.lerobot_info import get_sys_info, format_dict_for_markdown

# Get system info as dictionary
info = get_sys_info()

print(f"LeRobot version: {info['LeRobot version']}")
print(f"Python version: {info['Python version']}")
print(f"CUDA available: {info['Is PyTorch built with CUDA support?']}")

# Get formatted markdown string
markdown = format_dict_for_markdown(info)
print(markdown)

​
Custom Information Gathering

import platform
import importlib

def get_package_version(package_name: str) -> str:
    """Get package version or 'N/A' if not installed."""
    try:
        module = importlib.import_module(package_name)
        return getattr(module, "__version__", "Installed (version not found)")
    except ImportError:
        return "N/A"

# Check specific packages
print(f"Numpy: {get_package_version('numpy')}")
print(f"PyTorch: {get_package_version('torch')}")
print(f"Platform: {platform.platform()}")

​
Script Features

​
Minimal Dependencies

The script is designed to run with minimal dependencies:
  • Uses only Python standard library
  • Gracefully handles missing packages
  • Returns “N/A” for unavailable information
  • Can run even if LeRobot is not installed

​
Safe Execution

The script:
  • Never modifies system state
  • Only reads information
  • Handles errors gracefully
  • Works on Windows, Linux, and macOS

​
See Also

Build docs developers (and LLMs) love

Get started for freeTalk to us