Quick start: Build and deploy a machine learning model in a Jupyter notebook

You need a project to store the data and the AutoAI experiment. You can use an existing project or create a project.

1. From the Navigation Menu , choose Projects > View all projects

2. Open an existing project. If you want to use a new project:

   1. Click New project.

   2. Select Create an empty project.

   3. Enter a name and optional description for the project.

   4. Choose an existing object storage service instance or create a new one.

   5. Click Create.

3. When the project opens, click the Manage tab and select the Services and integrations page.

   To preview this task, watch the video beginning at 00:07.

   1. On the IBM services tab, click Associate service.

   2. Select your watsonx.ai Runtime instance. If you don't have a watsonx.ai Runtime service instance provisioned yet, follow these steps:

      1. Click New service.

      2. Select watsonx.ai Runtime.

      3. Click Create.

      4. Select the new service instance from the list.

   3. Click Associate service.

   4. If necessary, click Cancel to return to the Services & Integrations page.

For more information or to watch a video, see Creating a project.
For more information on associated services, see Adding associated services.

To preview this task, watch the video beginning at 00:18.

You will use a sample notebook in this tutorial. Follow these steps to add the sample notebook to your project:

1. Access the Use sckit-learn to recognize hand-written digits notebook in the Resource hub.

2. Click Add to project.

3. Select the project from the list, and click Add.

4. Verify the notebook name and description (optional).

5. Select a runtime environment for this notebook.

6. Click Create. Wait for the notebook editor to load.

7. From the menu, click Kernel > Restart & Clear Output, then confirm by clicking Restart and Clear All Outputs to clear the output from the last saved run.

To preview this task, watch the video beginning at 00:44.

The first section in the notebook sets up the environment by specifying your IBM Cloud credentials and watsonx.ai Runtime service instance location. Follow these steps to set up the environment in your notebook:

1. Scroll to the Set up the environment section.

2. Choose a method to obtain the API key and location.

   • Run the IBM Cloud CLI commands in the notebook from a command prompt.

   • Use the IBM Cloud console.

     1. Launch the API keys section in the IBM Cloud Console, and create an API key.

     2. Access your IBM Cloud resource list, view your watsonx.ai Runtime service instance, and note the Location.

     3. See the watsonx.ai Runtime API Docs for the correct endpoint URL. For example, Dallas is in us-south.

3. Paste your API key and location into cell 1.

4. Click the Run icon to run your code in cells 1 and 2.

5. Run cell 3 to install the ibm-watson-machine-learning package.

6. Run cell 4 to import the API client and create the API client instance using your credentials.

7. Run the cell with the code client.spaces.list(limit=10) to see a list of all existing deployment spaces. If you do not have a deployment space, then follow these steps:

   1. Open another tab with your Cloud Pak for Data deployment.

   2. From the Navigation Menu , click Deployments.

   3. Click New deployment space.

   4. Add a name and optional description for the deployment.

   5. Click Create, then View new space.

   6. Click the Manage tab.

   7. Copy the Space GUID and close the tab, this value will be your space_id.

8. Copy and paste the appropriate deployment space ID into the cell with the code space_id = 'PASTE YOUR SPACE ID HERE', then run that cell and the cell with the code client.set.default_space(space_id) to set the default space.

To preview this task, watch the video beginning at 02:14.

Now that all of the environment variables are set up, you can run the rest of the cells in the notebook. Follow these steps to read through the comments, run the cells, and review the output:

1. Run the cells in the Explore data section.

2. Run the cells in the Create a scikit-learn model section to.

   1. Prepare the data by splitting it into three data sets (train, test, and score).

   2. Create the pipeline.

   3. Train the model.

   4. Evaluate the model using the test data.

3. Run the cells in the Persist locally created scikit-learn model section to publish the model, get model details, and get all models.

   Note:

   If you are using Runtime 24.1 on Python 3.11, then you will need to change the software_spec_uid to runtime-24.1-py3.11 and the scikit-learn version to scikit-learn-1.3.

4. Run the cells in the Deploy and score section to create the online deployment, get deployment details, and send a scoring request to the deployed model to see the prediction.

5. Click File > Save.

To preview this task, watch the video beginning at 04:07.

You can also view the model deployment directly from the deployment space. Follow these steps to test the deployed model in the space.

1. From the Navigation Menu , click Deployments.

2. Click the Spaces tab.

3. Select the appropriate deployment space from the list.

4. Click Scikit model.

5. Click Deployment of scikit model.

6. Review the Endpoint and Code snippets.

7. Click the Test tab. You can test the deployed model by pasting the following JSON code:

      {"input_data": [{"values": [[0.0, 0.0, 5.0, 16.0, 16.0, 3.0, 0.0, 0.0, 0.0, 0.0, 9.0, 16.0, 7.0, 0.0, 0.0, 0.0, 0.0, 0.0, 12.0, 15.0, 2.0, 0.0, 0.0, 0.0, 0.0, 1.0, 15.0, 16.0, 15.0, 4.0, 0.0, 0.0, 0.0, 0.0, 9.0, 13.0, 16.0, 9.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 14.0, 12.0, 0.0, 0.0, 0.0, 0.0, 5.0, 12.0, 16.0, 8.0, 0.0, 0.0, 0.0, 0.0, 3.0, 15.0, 15.0, 1.0, 0.0, 0.0], [0.0, 0.0, 6.0, 16.0, 12.0, 1.0, 0.0, 0.0, 0.0, 0.0, 5.0, 16.0, 13.0, 10.0, 0.0, 0.0, 0.0, 0.0, 0.0, 5.0, 5.0, 15.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 8.0, 15.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 13.0, 13.0, 0.0, 0.0, 0.0, 0.0, 0.0, 6.0, 16.0, 9.0, 4.0, 1.0, 0.0, 0.0, 3.0, 16.0, 16.0, 16.0, 16.0, 10.0, 0.0, 0.0, 5.0, 16.0, 11.0, 9.0, 6.0, 2.0]]}]}
   

8. Click Predict. The resulting prediction indicates that the hand-written digits are 5 and 4.

If you'd like to remove all of the assets created by the notebook, create a new notebook based on the Machine Learning artifacts management notebook. A link to this notebook is also available in the Clean up section of the Use scikit-learn to recognize hand-written digits notebook used in this tutorial.