Exercise Session 1: Introduction to Machine Learning and Deep Learning

Connect to your HT Jupyter instance...

As we did likely in the connection tutorial, you need to:

1. SSH into our cluster to enable port forwarding. The command is something like:

ssh your.user@hpclogin.fht.org -L 8888:gnodeXX:YYYY

Where you obviously will have to replace XX and YYYY, as well as insert your real user name.

2. Now connect to your Jupyter instance from your local broser by going to:

localhost:8888

Clone this repo...

In Jupyter...

• Open a terminal window (inside the browser, from within Jupyter).
• Clone this repository by writing git clone https://github.com/dl4mia/01_intro_mldl.git.
• A new folder was created, containing all three exercises. In order to run them we need to first create a stutable conda environment.

Setup conda...

From within the same terminal in your browser, create a conda environment for this exercise and activate it:

conda create -n 01_intro_mldl python=3.8
conda activate 01_intro_mldl

Then install the following packages

conda install tensorflow-gpu keras jupyter matplotlib

Now navigate to the exercise folder we cloned just before and start with the exercises!