Deep Learning Course @ VU Amsterdam

Installation

NOTE works for MacOS and Linux - for Windows it might work in some setups, but was not tested

Clone this repo including submodules:

# with https (user + password) auth
git clone --recurse-submodules https://github.com/boczekbartek/deep_learning.git

# with ssh (ssh key based) auth:
git clone --recurse-submodules git@github.com:boczekbartek/deep_learning.git

Create conda or pip virtual env (conda is recommended)
```
conda create --name dl python=3.9
```
Install required packages
- If you are on machine with CUDA GPU (like Nvidia gpus) use:
```
./conda_install_cuda.sh
```
- Or if you don't have CUDA
```
./conda_install.sh
```

Assignment 1 - implementing gradient descent manually in numpy

Code directory: A1

Assignment 2 - Automatic differentiation

Code directory: A2

Assignment 3 - RNN or CNN

RNN

Code directory: A3-rnn

CNN:

Code directory: A3-cnn

Assignment 4 - Generative Models / Graph CNNs / Reinforcement Learning

VAE + GAN

Code directory: A4-vae-gan

Graph CNNs

Code directory: A4-graph-cnn

Deep Reinforcement Learning

Code directory: A3-reinforcement