For us to all be on the same page, we created a fake dataset that represents the data for a creditworthiness model for loan purposes. This dataset has “train”, “test” and “inference” data, along with "feedback" data (or “ground truth”) to check whether the loans were in fact paid back.
Monitoring your AI system is a good idea and can easily be achieved with Mona also if you don’t have ground-truth data for your inference runs.
In the first part of the tutorial, we will start with downloading the tutorial data set and SDK export scripts, getting export credentials (API key and secret), exporting the data to Mona and uploading an initial configuration file.
To begin our tutorial, let's first clone the Github repository to our environment.
You can clone the repo by clicking on the "code" tab and then clicking on the copy link button. Then in your terminal under your chosen directory type: "git clone [copied URL]"
Now let’s first take a look at a sample of our data so we know what we are working with.
Updated 23 days ago