samsi-deep-learning

Practical materials for the Deep Learning course at SAMSI/Duke-UNC-NC State, using the PyTorch DL framework.

Installation

• Make sure you have a working Python 3 environment. We recommend you to use Anaconda

• After that, you need to install PyTorch. This course is based on PyTorch 1.2, which is the most stable version at the time of this writing. You can install from here

Topics

1. An introduction to the pytorch framework. What does it have to offer compared to numpy etc? notebook

2. A basic logistic regression and multilayer perceptron for the MNIST dataset. notebook

3. Intro to NLP using torchtext, the IMDB dataset and both an LSTM network and a Transformer. notebook

4. Intro to adversarial examples, attacking a convolutional network. notebook

5. Intro to generative adversarial networks, using the MNIST dataset. notebook

Contact

• Should you have any questions or problems, open an issue or email Victor Gallego (victor.gallego@icmat.es) or Roi Naveiro (roi.naveiro@icmat.es).