flâneur

Torchvista: Building an Interactive Pytorch Visualization Package for Notebooks | Towards Data Science

towardsdatascience.com · 1,862 words · saved by 1 readers

We use cookies to enhance your browsing experience, serve personalised ads or content, and analyse our traffic. By clicking "Accept All", you consent to our use of cookies. No description available. Blueconic sets this cookie as a unique identifier for the BlueConic profile. No cookies to display. Publish AI, ML & data-science insights to a global community of data professionals. Building a tool to interactively visualize the forward pass of any Pytorch model from within notebooks. In this post, I talk through the motivation, complexities and implementation details of building torchvista, an open-source package to interactively visualize the forward pass of any Pytorch model from within web-based notebooks. To get a sense of the workings of torchvista while reading this post, you can check out: Github page if you want to install it via pip and use it from web-based notebooks (Jupyter, Colab, Kaggle, VSCode, etc) An interactive demo page with various well-known models visualized A Googl

Introduction In this post, I talk through the motivation, complexities and implementation details of building torchvista, an open-source package to interactively visualize the forward pass of any Pytorch model from within web-based notebooks. To get a sense of the workings of torchvista while reading this post, you can check out: Github page if you want to install it via pip and use it from web-based notebooks (Jupyter, Colab, Kaggle, VSCode, etc) An interactive demo page with various well-known models visualized A Google Colab tutorial A video demo: Motivation Pytorch models can get…

saved by

related reading