Netron is a cross-platform viewer for neural network, deep learning, and machine learning models, designed for developers and researchers working with various model formats. It provides a visual representation of model architectures, enabling easier understanding and debugging of complex networks.

How It Works

Netron parses and visualizes model files by analyzing their structure, layers, and operations. It supports a wide array of formats including ONNX, TensorFlow Lite, Core ML, Keras, PyTorch, and more, offering a unified interface for inspecting diverse model types.

Quick Start & Requirements

• Install:
  • macOS: brew install --cask netron
  • Linux: snap install netron
  • Windows: winget install -s winget netron
  • Python: pip install netron then run netron [FILE] or netron.start('[FILE]')
  • Browser: Available online.
• Prerequisites: None explicitly stated beyond OS-specific installers or Python.
• Resources: No specific hardware requirements mentioned.
• Links: Sample Models

Highlighted Details

• Supports over 15 model formats including ONNX, TensorFlow Lite, Core ML, Keras, and PyTorch.
• Offers experimental support for formats like TorchScript, TensorFlow, and scikit-learn.
• Available as desktop applications, a browser version, and a Python package.

Maintenance & Community

• Actively maintained by Lutz Roeder.
• No specific community links (Discord/Slack) or roadmap are provided in the README.

Licensing & Compatibility

• MIT License.
• Permissive license allows for commercial use and integration into closed-source projects.

Limitations & Caveats

The README mentions "experimental support" for several formats, indicating potential instability or incomplete feature coverage for those specific types.