PyTorch Tutorial for Deep Learning

Explore the fundamentals of deep learning with PyTorch in this comprehensive 2-hour tutorial presented by Antoine Prouvost from Polytechnique Montréal, Canada at GERAD Research Center. Learn how to implement neural networks while receiving theoretical reminders, practical tips, and suggested code workflows. Discover PyTorch's capabilities as a simplified version of NumPy, featuring GPU acceleration and automatic gradient computation. Gain hands-on experience in efficiently training neural networks with minimal code overhead, including a practical exercise in developing a convolutional model for digit recognition. Cover essential topics such as tensors, GPU utilization, automatic differentiation, and dynamic neural networks. This tutorial is ideal for those with a background in Python, NumPy, basic machine learning concepts, and a fundamental understanding of neural networks, including feed-forward processes, backpropagation, and gradient descent.

PyTorch tutorial for deep learning, Antoine Prouvost