{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/46ccda9605af4cbfb8fecc62764062d8\" frameborder=\"0\" width=\"1108\" height=\"831\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":831,"width":1108,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":831,"thumbnail_width":1108,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/46ccda9605af4cbfb8fecc62764062d8-19947a5a1ab53659.gif","duration":4012.787,"title":"Colab Demo - Using GitHub and Google Colab for AI with FASHION-MNIST Data","description":"This Loom explains how to use GitHub and Google Colab together to run a real AI demo for educators. It describes GitHub as a shared filing cabinet where teachers can publish a notebook for others to download, and Colab as the cloud workbench that runs computation on Google’s machines, making GPUs accessible even on Chromebooks. The walkthrough focuses on an image classification notebook using Fashion MNIST, including checking GPU availability, setting runtime to a T4 GPU, importing tools and data, preprocessing pixels by dividing by 255, training on 60,000 images, and testing on 10,000 unseen images to monitor accuracy and overfitting. It also covers a parameter experiment by increasing hidden units and rerunning for results, then saving the final notebook to GitHub so it can be reopened later."}