Exercise - Build the model

  • 15 minutes

Now let's build our model! The first step is to download the dataset and create our Azure service.

Activate the sandbox

The process of signing in to activate the sandbox runs outside the learning module. You're automatically returned to the module after you sign in.

To activate the sandbox:

  1. Select Sign in to activate Sandbox. Enter your credentials to authenticate.

  2. If prompted, select Review permissions.

    Screenshot that shows the sandbox, with the Review Permissions button highlighted.

  3. Verify the permissions settings, and select Accept.

    Screenshot that shows the sandbox, with permissions details and the Accept button highlighted.

The message Sandbox activated! appears, and you can continue with the learning module.

Download the dataset

Data is the first thing we need to create a machine learning model. We'll use a subset of the NABirds dataset from the Cornell Lab to train our model.

Download the zip file that contains the dataset:

  1. In your web browser, go to the dataset in GitHub.

  2. Select Download.

    Screenshot that shows GitHub in a web browser, with the dataset zip file and the Download button highlighted.

    The zip file is copied to your computer and stored in your default location for downloaded files.

  3. When the download is finished, unzip the file. Note the folder location because you’ll need it in a later step.

Create a Custom Vision API resource

Next, we'll create an API resource in Azure AI Custom Vision.

  1. In the Azure portal, select Create a resource.

  2. Search for custom vision. In the Custom Vision card in the search results, select Create.

    Screenshot that shows searching for Custom Vision resource templates in the Azure portal.

  3. On the Basics tab, enter or select the required values:

    1. Select your Azure subscription.

    2. Create a new resource group:

      1. Beside Resource group, select the Create new link.

      2. In the dialog box, enter BirdResourceGroup, and then select OK.

      Screenshot that shows how to create a new resource group in the Azure portal.

    3. Enter a name for your new Custom Vision service resource (for example, BirdCustomVisionService).

    4. Under Instance Details:

      1. Set Region to (US) South Central US.

      2. Enter a Name for your resource, using only alphanumeric characters and hyphens.

    5. Under Training Resource, set Training pricing tier to Free F0 (2 Transactions per second…).

    6. Under Prediction Resource, set Prediction pricing tier to Free F0 (2 Transactions per second…).

  4. Select Review + create.

  5. Select Create.

    Screenshot that highlights the elements to select in the Azure portal to create a new Custom Vision resource.

When the deployment finishes, select Go to resource group.

Screenshot that shows the deployment finished page in the Azure portal, with the Go to resource button highlighted.

You should see two resources listed in this new resource group: a training resource, and a prediction resource.