gregunz / DeepLearning2019

Deep Learning (EE-559 @EPFL) Spring 2019 - Projects

Deep Learning Projects

Deep Learning (EE-559 @EPFL) Spring 2019

Project 1 - Weight Sharing and Auxiliary Loss

cd Proj1

How to run the code

Dependencies

Following libraries are required:

matplotlib
pandas
pickle
torch
torchvision
seaborn

They can be installed using the following command:

pip install <insert lib here>

Test

In order to test the results claimed in the report, please run the following test.py file using the following command:

python3 test.py

One can use its own parameters either by modifying the constants.py file or by using the following command line arguments:

usage: test.py [-h] [--models MODELS] [--epochs EPOCHS] [--rounds ROUNDS]
               [--seed SEED] [--aux AUX_LOSS_FACTOR]

Project 2 - Deepy

cd Proj2

Dependencies

Following libraries are required:

torch

They can be installed using the following command:

pip install <insert lib here>

Example

An simple example is provided in the report and a full example is provided in the file test.py which can be used with the following command:.

python3 test.py