Intro to deep learning workshop.

You might be working on other Python projects at the moment that use some of the libraries from the workshop - but at different versions. To avoid breaking your current setup - and to simplify the installation process - we will use Conda environments.

The idea is to create a dl-workshop environment for the workshop that you can activate when you work on it and then deactivate to restore your original configuration.

We will use Conda to install the Python libraries for the workshop. You can get it by installing Anaconda (Python 3.6 version).

Verify that it's installed by running the following command in a terminal (on macOS or Linux) or the "Anaconda Prompt" on Windows. The command should return its version number.

conda -V

Make sure that you have a recent version (see changelog). To update Conda, navigate to its folder (usually in your home directory) and run

conda update conda

If you correctly installed Conda, you should be able to list the current environments by running this command in a terminal (or the Anaconda Prompt for Windows users).

conda env list

Let's create a dl-workshop entry for the workshop. Start by downloading this GitHub repository ("download" button from above) and extract it on your computer. The folder contains .yml configuration files that list the libraries needed for the workshop with their version number. Use the file that matches your setup - Windows, macOS or Linux.

To create the environment, open your terminal and navigate to the folder with the .yml files. You can then install the environment by typing conda create -f .. with the appropriate file.

# For Windows users
conda env create -f environment-windows.yml

# On macOS
conda env create -f environment-macos.yml

# .. Linux
conda env create -f environment-linux.yml

This operation might take some time. Verify that the environment is installed with conda env list.

You can activate the environment with

# On Windows
activate dl-workshop

# On macOS and Linux
source activate dl-workshop

Troubleshooting: If this doesn't work for you, then it's likely that Conda doesn't find the environment. Is it installed?

Now, everything that you do from this terminal uses the libraries from the dl-workshop environment. To launch the Jupyter interface, type

# Open the Jupyter dashboard
jupyter notebook

# Or the new "Jupyter lab" one
jupyter lab

# You can then quit the interface with Ctrl+C (twice on Windows)

Similarly, you can deactivate the environment with

# On Windows
deactivate

# On macOS and Linux
source deactivate

Try to run the scavenger-hunt.ipynb notebook (and the other ones!) with the workshop environment.