Netron

Netron is a viewer for neural network, deep learning, and machine learning models, supporting a wide range of formats across platforms including web, desktop, and Python.

🚀 Overview

Netron provides an intuitive way to visualize complex machine learning model architectures, solving the problem of opaque 'black box' models by rendering interactive graphs of layers, nodes, tensors, and attributes.

It is built for machine learning engineers, data scientists, AI researchers, and developers who need to inspect, debug, understand, or present neural network structures without framework-specific tools.

Created by Lutz Roeder, a Principal Software Engineer at Microsoft, Netron was developed to offer universal support for popular model formats, making it easier to explore models from various ecosystems like ONNX, PyTorch, TensorFlow, and more.

✨ Key Features

Interactive model visualization with support for stable formats including ONNX, TensorFlow Lite, PyTorch, torch.export, ExecuTorch, Core ML, Keras, Caffe, Darknet, TensorFlow.js, Safetensors, and NumPy
Experimental support for TorchScript, MLIR, TensorFlow, OpenVINO, RKNN, ncnn, MNN, PaddlePaddle, GGUF, scikit-learn, and recent additions like JAX
Cross-platform: browser-based app at netron.app, native desktop apps for Windows, macOS, Linux, and Python integration via pip
Sample models available for quick testing