<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Experiments on George Chemmala</title><link>https://george.chemmala.com/experiments/</link><description>Recent content in Experiments on George Chemmala</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Mon, 26 May 2025 23:44:59 -0400</lastBuildDate><atom:link href="https://george.chemmala.com/experiments/index.xml" rel="self" type="application/rss+xml"/><item><title>Track Active Users</title><link>https://george.chemmala.com/experiments/track_active_users/</link><pubDate>Mon, 26 May 2025 23:44:59 -0400</pubDate><guid>https://george.chemmala.com/experiments/track_active_users/</guid><description>An experiment in using a serverless cloud function</description></item><item><title/><link>https://george.chemmala.com/experiments/mnist/mnist/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://george.chemmala.com/experiments/mnist/mnist/</guid><description>&lt;!DOCTYPE html> MNIST Example Here is a simple TensorFlow.js example that loads the MNIST dataset, trains a model, and allows you to draw digits to see how well the model predicts them. Once the model is trained, you can draw a digit in the box below and it will predict what digit you drew. The model is trained on the MNIST dataset, which consists of 28x28 pixel grayscale images of handwritten digits (0-9).</description></item><item><title/><link>https://george.chemmala.com/experiments/mnist/readme/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://george.chemmala.com/experiments/mnist/readme/</guid><description>MNIST Tensorflow.js Demo # You can try out the demo here:
MNIST TensorFlow.js Demo
This demo uses TensorFlow.js to load the MNIST dataset, train a model, and allow you to draw digits to see how well the model predicts them. Once the model is trained, you can draw a digit in the box below and it will predict what digit you drew. The model is trained on the MNIST dataset, which consists of 28x28 pixel grayscale images of handwritten digits (0-9).</description></item></channel></rss>