gdario/amld-pytorch-workshop

Introduction to PyTorch Workshop at the AMLD 2019

This is a tutorial presented during a workshop at the Applied Machine Learning Days 2019.

Binder

Google Colab

As an alternative to Binder, you can also use Google Colaboratory, though you should use Binder if possible.

Using Google Colab

The Google Colab notebooks are available under:

• 1-Basics.ipynb
• 2-Autograd.ipynb
• 3-Optimization.ipynb
• 4-Modules.ipynb
• 5-CNN-LSTM.ipynb
• 6-Transfer-Learning.ipynb

In order to use Google Colab, you have to login using your Google account:

Changing the runtime type

You can add GPU support on Google Colab by changing the runtime type as depicted below:

During the Workshop

During the workshop, we highly recommend to use Binder or Google Colab. If you want to run the notebooks again later, you can use the following setup using Anaconda. Unfortunately, we won't have time to help you with your conda installation.

Using conda

If you want to run the notebooks locally, you can use conda. The following instructions should work on Linux/Mac OS, Windows might require slight adaptations.

Step 1: Install conda

If you have not installed it yet, you can download it from Anaconda (Python 3.6 version).

Verify that it is installed by running

conda -V

Make sure your conda installation is up-to-date:

conda update conda

Step 2: Download repository and install environment

Now clone the repository:

git clone https://github.com/ahug/amld-pytorch-workshop.git
cd amld-pytorch-workshop

The available conda environments can be listed using

conda env list

Let's now create a new environment called 'amld-pytorch'.

conda env create -f environments.yml

Step 3: Activate/Deactivate the environment

After the environment has been created, you can activate it by

source activate amld-pytorch

Now start the Jupyter notebook by running

jupyter notebook

The environment can similarly deactivated by

source deactivate

Additionnal resources:

Check out these others tutorials and courses:

• Official tutorials : https://pytorch.org/tutorials/
• PyTorch for DL (.py files): https://github.com/yunjey/pytorch-tutorial
• PyTorch for DL (notebooks): https://github.com/yandexdataschool/Practical_DL
• PyTorch for Numpy users : https://github.com/wkentaro/pytorch-for-numpy-users