Summary
This chapter introduced you to an open source alternative for creating an AutoML process using the AutoGluon Python library. We also used AutoGluon's Tabular predictor to advance the Age Calculator use case and demonstrated how to find the best-suited model for the tabular dataset.
We further expanded on the AutoML methodology to address a complicated computer vision use case by finding the best-suited CNN model for the Rock Paper Scissors dataset. This was accomplished using AutoGluon's Image predictor and further optimized using SageMaker's GPU-based ML instances. This chapter also introduced the concept of a runtime process artifact, in the form of a container image.
In the next chapter, we will continue to expound on this concept and introduce how an ML runtime artifact can further streamline the ML process, especially when the artifact is used in conjunction with other AWS services.
