What is torch-video-tutorials?

This repo is a curated set of video tutorials that teach you how to use Torch, a framework for machine learning and deep learning. It is designed to make Torch less intimidating for newcomers by walking you through the basics with videos, slides, and quizzes. The content is organized into a progressive course. It starts with fundamentals like Lua (the programming language Torch is built on) and how to work with data structures called Tensors. From there, it moves into artificial neural networks, explaining how they make predictions and how they learn from data. Later sections cover convolutional neural networks (used heavily for image tasks) and recurrent neural networks (used for sequences like text or time series), with each topic pairing a video walkthrough with slides and sometimes code scripts or quizzes. This is for someone who wants to learn deep learning concepts hands-on but prefers a guided, visual approach rather than reading dense documentation. A student exploring computer vision, for example, could watch the CNN section and follow along with the provided Lua scripts to build and train models like LeNet or AlexNet. A self-taught practitioner trying to understand how backpropagation works could use the neural network videos and quizzes to test their knowledge. A notable thing about this project is that Torch itself is an older framework that has largely been superseded by newer tools, and Lua is not a mainstream language in modern machine learning. The tutorials still hold educational value for understanding core deep learning mechanics, but someone looking to build production systems today would likely need to translate these concepts to a more current framework.