tensorViz Model Explorer

tensorViz Model Explorer offers a web-based interface to interactively explore and understand a curated set of 29 PyTorch model architectures. Users can visually inspect tensor shapes, layer structures, and operations with step-by-step guided lessons covering attention mechanisms, vision models, language models, generative models, and more. It provides runnable source examples and the ability to integrate exploration with VS Code. No software installation or Python environment setup is required.