Experiment analytics with SageMaker Experiments
In the previous recipe, we ran multiple training jobs and used SageMaker Experiments to keep track of the parameters, input and output artifacts, metric values, and other metadata with the Experiment, Trial, TrialComponent, and Tracker resources using the smexperiments library.
In this recipe, we will use ExperimentAnalytics from sagemaker.analytics to load and analyze the DataFrame containing the details of the previous experiments we have performed and tracked using SageMaker Experiments. This allows us to inspect and analyze the results of the training jobs with just a few lines of code.
Getting ready
The following is the prerequisite for this recipe:
- This recipe continues from Running and managing multiple experiments with SageMaker Experiments.
How to do it…
The first set of steps in this recipe focuses on preparing and loading the prerequisites:
- Navigate to the
my-experiments/chapter05directory...
