• Hands-On Introduction to PyTorch

    AIR - Academy of Innovation & Research

    Neural networks have played a critical role in recent advances in automation and artificial intelligence over the last decade - with applications in sectors as broad as medicine, driver-less cars and industrial process optimisation. PyTorch and TensorFlow are the two leading machine learning frameworks. PyTorch is more pythonic in design, and so much easier to use and debug. --- In this beginner-friendly tutorial we will: * learn the basics of how a neural network works (without too much maths) * learn the basics of PyTorch * see the basic structure of PyTorch machine learning code * build step-by-step a neural network with PyTorch that learns to classify images of hand-written digits, a standard machine learning challenge * build the confidence to continue to learn about machine learning, and experiment with PyTorch --- This is a session designed for beginners to both machine learning and PyTorch. You don't need to be an expert with coding and Python but a little experience with be helpful. --- This is a hands-on practical tutorial. Please bring a fully charged laptop (access to power sockets not guaranteed), with working wifi and a modern browser like Chrome, Firefox or Safari. There is no need to install Python, we will use the online Google Colab service. You'll need a Google Account for this, and if you log into Gmail or YouTube you already have one.