From the course: Machine Learning with SageMaker by Pearson

Unlock this course with a free trial

Join today to access over 25,200 courses taught by industry experts.

Hyperparameter tuning demonstration

Hyperparameter tuning demonstration - Amazon SageMaker Tutorial

From the course: Machine Learning with SageMaker by Pearson

Start my 1-month free trial Buy for my team

Hyperparameter tuning demonstration

“

fine-tune the quality of the results that we get from our models. And SageMaker has the ability to do automated model tuning, wherein you define training jobs and it will execute them, and it will increment a hyperparameter had the best metric. And that's the one you would then put into production. OK, let's create a training job definition here. We're going to use a built-in algorithm. And let's use XGBoost. Input mode is going to come from S3 file. And then here are those objective metrics that we can then try to optimize. So I just mentioned validation log loss. And that's what we're going to do in this demo. The problem is, and that bug that I mentioned a moment ago, that this has never worked for me, is that as we move forward with selection of options within here. We specify the channel name of train or validate. Input mode, we'll do a file content type, say text CSV. No compression, no wrapper S3. Fully replicated, we give it a path to our data. Checkpoint config and then…

Contents