Summary

In this chapter, you got an overview of the various ways you can create an ML model in the AzureML workspace. You started with a simple regression model that was trained within the Jupyter notebook's kernel process. You learned how you can keep track of the metrics from the models you train. Then, you scaled the training process into the cpu-sm-cluster compute cluster you created in Chapter 7, The AzureML Python SDK. While scaling out to a remote compute cluster, you learned what the AzureML environments are and how you can troubleshoot remote executions by looking at the logs.

In the next chapter, you will build on this knowledge and use multiple computer nodes to perform a parallelized hyperparameter tuning process, which will locate the best parameters for your model. You will also learn how you can completely automate the model selection, training, and tuning using the AutoML capabilities of the AzureML SDK.