Sagemaker Xgboost Framework 1.7.0+ Support?

0

Any plans for Built-in Sagemaker Xgboost to support version 1.7.0+? This version allows for spark based xgboost models.

已提问 2 年前345 查看次数
1 回答
-1

Hi,

At this time sagemaker does support Xgboost Framework 1.7.0+

The current support versions of XGBoost Algorithm on Sagemaker is the following:

Framework (open source) mode: 1.0-1, 1.2-1, 1.2-2, 1.3-1, 1.5-1

Algorithm mode: 1.0-1, 1.2-1, 1.2-2, 1.3-1, 1.5-1

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

However, you still can create a create a custom Create a container that can have support XGBoost 1.7.2, and can host your algorithms and models.

For more details on creating your own container please refer to "Create a container with your own algorithms and models" in the link below.

https://docs.aws.amazon.com/sagemaker/latest/dg/docker-containers-create.html

In addition there are also some examples that you can also use to help you get started on creating your own container.

Build a Custom Training Container and Debug Training Jobs with Amazon SageMaker Debugger - https://sagemaker-examples.readthedocs.io/en/latest/sagemaker-debugger/build_your_own_container_with_debugger/debugger_byoc.html
Example Notebooks: Use Your Own Algorithm or Model - https://docs.aws.amazon.com/sagemaker/latest/dg/docker-containers-notebooks.html
AWS
Chris_T
已回答 2 年前
  • I must be confused, the link you provided states:

    The current release of SageMaker XGBoost is based on the original XGBoost versions 1.0, 1.2, 1.3, and 1.5.

    How would I specify a built-in algorithm to use in hyperparameter tuning leveraging XGBoost version 1.7?

    The documentation for image URI also doesn't seem to specify 1.7: https://docs.aws.amazon.com/sagemaker/latest/dg/ecr-us-west-1.html#xgboost-us-west-1.title

您未登录。 登录 发布回答。

一个好的回答可以清楚地解答问题和提供建设性反馈,并能促进提问者的职业发展。

回答问题的准则