CODEWITHSUNDEEP
xgboostamazon-sagemaker
Alex
Alex

Reputation: 320

Sagemaker XGBoost Version 1.7?

I am trying to leverage Sagemaker's built-in algorithms for training/prediction. I have hyperparameter tuning and batch transform running, but I wanted to test the spark distributed XGBoost model version which seems to be available in version 1.7+. As far as I can tell from the documentation:

https://docs.aws.amazon.com/sagemaker/latest/dg/xgboost.html#xgboost-supported-versions

This is NOT a supported version for Sagemaker? However, I received an incomplete answer on AWS forums that said the above link shows 1.7 IS a supported version.... Anyone able to clarify for me? I am not able to load a SM trained XGBoost model into the Spark distributed version.

Upvotes: 0

Views: 638

Answers (2)

zdarova
zdarova

Reputation: 152

Currently (April 26, 2023) Amazon SageMaker SDK supports XGBoost 1.7.

The documentation: https://docs.aws.amazon.com/sagemaker/latest/dg/xgboost.html

Upvotes: 1

Kirit Thadaka
Kirit Thadaka

Reputation: 507

As shown in the documentation, 1.7 is not supported in SageMaker's native XGBoost estimator. If you require 1.7, it would be best to use your own container and run your jobs through an Estimator with your custom container.

Upvotes: 1

Related Questions