Model Explorer is a powerful graph visualization tool designed to help you understand and debug your ML models. Key features include:

• Hierarchical Layout

  View your model's structure clearly with nested layers that you can expand and collapse as needed.

• Metadata Overlays

  Gain insights by overlaying metadata (e.g., attributes, inputs/outputs, etc) and custom data (e.g. performance) on nodes.

• Powerful Interactive Features

  Focus on specific areas with search, navigate graphs smoothly with bookmarks and layer popups, customize node styles with queries, compare graphs side-by-side, and more.

• Popular ML Framework Support

  Works with popular ML formats like TensorFlow, JAX, TFLite, TensorFlow.js, MLIR, and PyTorch.

• Extension Support

  Write your own extensions in Python to add support for more model formats.

• High Performance Rendering

  Model Explorer uses custom WebGL shaders for GPU-accelerated rendering, effortlessly visualizing tens of thousands of nodes, edges, and labels, and ensuring a constant 60 FPS for responsive user interaction.