Learn PyTorch and implement deep neural networks (and classic machine learning models). This is a hands on tutorial which is geared toward people who are new to PyTorch.
PyTorch is a relatively new neural network library which offers a nice tensor library, automatic differentiation for gradient descent, strong and easy gpu support, dynamic neural networks, and is easy to debug.
We will cover: - Machine Learning 101 recap: model + loss + optimization - PyTorch basics - tensors (and variables) - automatic differentiation and gradient descent - PyTorch's take on neural networks - Deep neural networks / convolutional networks for computer vision - transfer learning and fine-tuning - build your convolutional network from scratch
If there is time left we might touch on: - Implement a Recurrent Neural Networks (RNN) from scratch - Simple recommender engines/collaborative filtering
Goals: - understand PyTorch's concepts - be able to use transfer learning in PyTorch - build simple PyTorch models from scratch
Prerequisites: - you have implemented machine learning models yourself - you know what deep learning is - you have used numpy - maybe you have used tensorflow or similar libs - if you use PyTorch on a daily basis, this tutorial is probably not for you
Materials will be made available via https://github.com/sotte/pytorch_tutorial. Please install the dependencies as instructed in the README.md.