willyfh/visualtorch

VisualTorch aims to help visualize Torch-based neural network architectures.

What it solves

VisualTorch provides a way to visualize PyTorch neural network architectures, making it easier for developers and researchers to understand, document, and present their model structures. It supports both simple sequential models and more complex custom models.

How it works

The tool uses a tracing-based approach, performing a real forward pass with dummy input to map out the model's architecture. It can then render these architectures in three distinct visual styles: flow-style, graph-style, and LeNet-style. Additionally, it can generate animated GIF reveals that show the model being built layer by layer.

Who it’s for

It is designed for PyTorch users, including researchers and developers who need to create professional diagrams of their neural networks for publications, presentations, or internal documentation.

Highlights

  • Multiple Visualization Styles: Supports flow, graph, and LeNet-style diagrams.
  • Animated Reveals: Ability to render the model architecture as an animated GIF.
  • MCP Integration: Includes an optional MCP server for generating diagrams from model source via stdio.
  • Research-Backed: Used in published research across Nature, IEEE, and MDPI.