2022-11-28
Enhancements
Drift Monitors: Integer Binning
A more intuitive understanding of model drift with enhanced bin ranges for data sent in as integers. This is enabled automatically for integer data type columns.
GraphQL API Update
ThevalidationResult object returned from createFileImportJob when the dryRun parameter is set to true has been changed to return an error type which includes the error code, the error message, and the error row. Additionally, the errorFilePath field has been renamed to filePath.In The News
How To Monitor Ranking Models
A primer on relevant model metrics and best practices for tackling model performance degradation.
Shipping Your Image Classification Model With Confidence
Dive into a code-along blog for image classification models! Learn how to develop an image classification model, prepare and ingest embedding data, and analyze embedding drift in this action-packed piece that mirrors the Colab.
Introducing Xander Song, Arize’s New Developer Advocate
Nine questions with Arize’s new Developer Advocate.
What is Recall?
Check out the latest in our series on model performance metrics! This practical primer explains where recall is most useful and the intuition behind the metrics.