Persontyle Workshop for Applied Deep Learning

by Amaia Salvador and Santiago Pascual (Universitat Politecnica de Catalunya, 2017)

This repository contains several ipython notebooks with machine learning tutorials using python, keras and tensorflow.

Contents

1. Installation
2. Setup
3. Sessions
4. Usage

Installation

• This code has been tested in a Linux machine with python 2.7. It should work on Mac OS X as well.
• Follow these instructions to install tensorflow 0.10 with pip.
• Install other dependencies pip install -r requirements.txt.

To run this code on GPU, you will need cuda and cudnn installed in your machine. Alternatively, you can follow these instructions to setup an AWS instance with jupyter notebook.

Setup

Run python setup.py to collect datasets and models before we begin, to avoid problems if notebooks freeze at some point.

Sessions

• AWS Setup
• Logistic Regression
• Multiclass Classification
• Deep models
• Convolutional Neural Networks
• Visualization
• GAN
• Deep Dream
• Neural Style
• Object Detection I
• Object Detection II
• Object Tracking I
• Object Tracking II
• LSTM char generation
• Embeddings
• 20news
• Question Answering

Usage

Run a notebook

jupyter notebook

Then navigate to:

http://localhost:8888

and start editing the notebooks.